Action Assembly Theory

John O Greene. Encyclopedia of Communication Theory. Editor: Stephen W Littlejohn & Karen A Foss. 2009. Sage Publication.

Action assembly theory (AAT) is an approach to explicating the processes by which people produce verbal and nonverbal messages. The domain encompassed by verbal and nonverbal message production is obviously quite broad, and thus AAT addresses issues such as the nature of consciousness, the processes that give rise to creativity in what people think and do, the link between thoughts and overt actions, the relationship between verbal and nonverbal components of behavior, and how people plan and edit what they say. AAT has been applied in investigating a range of phenomena, including the nature of the self-concept, the behavioral cues that accompany deception, communication skill and skill acquisition, communication apprehension, and conditions that affect speech fluency. The theory is most closely associated with John Greene, his colleagues, and his students.

Every theory is developed within a matrix of assumptions, methods, research findings, and even other theories. It is useful, then, to examine key elements of the intellectual matrix in which a theory is embedded. In the case of AAT, four hierarchically ordered, foundational influences are particularly noteworthy. At the most basic level, AAT reflects the commitments of science as a way of knowing (i.e., an emphasis on empiricism, inter-subjectivity, and rigor). Moving up one level, the particular branch of science reflected in the theory is that of cognitivism—explanation of behavior by recourse to descriptions of the mental states and processes that give rise to that behavior. Cognitive science itself encompasses a number of distinct philosophical and methodological approaches. The cognitive approach exemplified in AAT is primarily that of functionalism (i.e., inferring the nature of the mind from observed input-output regularities). At the fourth level of the hierarchy, the theory reflects the commitments of generative realism: the idea that people are simultaneously social, psychological, and physical beings and that theories of human behavior need to incorporate all three of these elements.

Just as one’s grasp of a theory is enhanced by understanding the assumptions and scholarly traditions that it reflects, it is similarly useful to be able to locate a theory along a timeline of pertinent intellectual developments. Relevant to AAT, cognitive science emerged as the dominant approach in experimental psychology in the mid-1950s, partly as a result of growing recognition of the inadequacies of behaviorism. The assumptions and techniques of cognitivism were quickly assimilated by scholars studying linguistics and speech production, but cognitive science was rather slow to gain a foothold in the field of communication proper. Not until the mid-1970s did scholars in the field begin to adopt a cognitive perspective in exploring communication processes. By the early 1980s, cognitivism had become an important force in the field, but, with a few notable exceptions, this work was focused almost exclusively on input processing (e.g., message comprehension) rather than output processes (e.g., message production). In this context, in 1984, the first article on AAT appeared—a publication that subsequently received the National Communication Association’s Woolbert Award for seminal contributions to communication research. In 1997, “A Second Generation Action Assembly Theory” (AAT2) was published, and research and conceptual refinements in the AAT framework continue to the present.

Key Aspects of the Theory

Pattern and Creativity

The particular form and substance of AAT was shaped by the intellectual traditions and assumptions described in the introduction above, but the single most important idea from which the theory sprang is a very simple observation that had been around long before AAT was developed. This fundamental idea is that all human behavior is both patterned and creative. In other words, a person’s behavior exhibits his or her characteristic ways of doing things—the topics the person talks about, what he or she has to say about them, the vocabulary used to say those things, ways of pronouncing those words, facial expressions and gestures that accompany his or her speech, and so on. At the same time that speech and nonverbal behavior reflect each person’s patterned ways of doing things, they are also always novel or creative: It turns out that we never exactly repeat ourselves, and even more importantly, we can use our patterned ways of thinking, speaking, and moving to create ideas, utterances, and actions that we have never produced before. In that simple observation lies the seed from which AAT grew: If behavior is both patterned and creative, then how can we understand where the patterned aspect of action comes from, and how can we understand how patterns give rise to novelty?

Mental Structures and Processes

As noted above, in keeping with a cognitive functionalist perspective, AAT attempts to explain behavior by specifying the nature of the mental structures and processes that give rise to that behavior. In the case of AAT, the fundamental question, again, is how to understand the patterned and creative property of human behavior. According to the theory, the patterned aspect of action arises from structures in long-term memory that have been acquired over the course of one’s lifetime. These memory structures, termed procedural records, are like tiny packets of information about what to do (i.e., action features)in particular situations in order to accomplish specific objectives. So a person might have learned, for example, to say “I’m sorry” in an effort to undo the damage in a situation in which he or she hurt another’s feelings.

Three things about procedural records are important to note. First, the memory codes in which they are represented reflect a hierarchy of levels of abstraction. Some codes, such as the example of saying “I’m sorry,” are abstract (conceptual, languagelike). But other procedural records are expressed in codes that are far more basic than that. Some records, for example, consist of the motor-code memories that a person has acquired for walking, reaching and grasping, pronouncing the sound units of his or her native language, and swinging a golf club. An extension of the first point, the second is that a single procedural record does not represent all the information used to produce a behavior. People do not, for example, have one record that holds all the information they use to flick on a light switch or raise a glass of water to their lips; rather, even simple movements such as these reflect the combination of multiple action features. And in the same way, a person may have an abstract record that indicates the need to say “I’m sorry,” but to actually produce that utterance, the person will have to rely on lower-level records for pronouncing those words. The third point is that people possess a very large number of procedural records. A conservative estimate might be that an adult possesses perhaps tens to hundreds of thousands of them.

In simplest terms, then, AAT holds that the patterned properties of behavior reflect the contents of procedural records held in long-term memory: Our behavior tends to reflect our own repertoire of ways of doing things. The creative character of behavior arises as a result of two processes by which the contents of procedural records are actually used in producing action. The first of these, activation, is basically a selection process that serves to retrieve those records that are relevant to one’s goals and the situation, and the second, assembly, then integrates or combines activated action features to produce one’s unfolding behavior. In essence, we think and do new things when we assemble new configurations of action features. Just as a child might construct an endless variety of forts, towers, and walls out of the same small set of building blocks, communicators constantly are assembling new thoughts, utterances, and nonverbal behaviors by combining novel configurations of action features.

The linchpin of AAT is the nature of the assembly process, which is described as coalition formation—combining action features that “fit” together, as, for example, when a high-level feature like the abstract notion to turn left meshes with motor-level features for turning the steering wheel and pressing the brake pedal. Activated action features that do not find their way into coalitions quickly decay and are not manifested in overt behavior. In contrast, those features that do mesh with others (a) stay activated longer, (b) are more likely to actually emerge in what a person says and does, and (c) are more likely to enter conscious awareness. The property of conscious awareness, in turn, brings to bear self-regulatory processes such as rehearsing, planning, problem solving, and editing.

The Nature of Message Behavior

From the perspective of AAT, human behavior is a complex, dynamic constellation of action features reflecting various hierarchically ordered representational codes. At any moment, only some coalitions will recruit motor-level features that permit them to be manifested in overt action; as a result, verbal and nonverbal behaviors reveal less than a person “means” (i.e., his or her momentary collection of coalitions). Conversely, because only a subset of coalitions is available to consciousness at any time, those same message behaviors will reveal more than the person “knows” (i.e., the contents of phenomenal awareness). In contrast to standard goals-plans-action models of message behavior, AAT presents a picture of a much more rapid, chaotic, and disjointed system underlying message production. AAT is also distinguished from models of social skill that suggest that skilled behavior is the product of motivation and ability. AAT holds that people may be motivated to behave in socially appropriate and effective ways, and also possess requisite knowledge for doing so, and yet fail to act in an optimal way because of the nature of the activation and assembly processes.

Assembly Difficulties and Failures

A key point of emphasis in the various empirical and theoretical applications of AAT has been on situations in which people encounter difficulties in assembling action features. As might be expected, one effect of assembly problems is to slow message production. As a result, studies of the time course of message behavior have been an important part of the AAT research program. For example, several experiments have focused on the production of messages designed to accomplish multiple social goals. The general finding of these studies is that when one’s social goals are incompatible (e.g., conveying a negative performance appraisal while showing support for the other), speech fluency is reduced. Another program of research, focused on skill acquisition, has examined patterns of improvement in speech fluency as a result of practice. An ongoing research program examines creative facility—individual differences in people’s ability to produce novel messages—in an effort to understand why some seem to be better than others at “thinking on their feet.”