A Theory of Action Identification

Robin R. Vallacher, Daniel M. Wegner

• Publisher: Psychology Press
• Year: 2014
• ISBN:
  9781317767862
• Edition: 1
• Topic: Psychologie
• Subtopic: Geschichte & Theorie in der Psychologie

First published in 1985. A person may be caught in the midst of a patently ridiculous act, interrupted in a moment of apparent confusion, or even aroused from sleep, and yet respond to a query of What are you doing? with remarkable ease. The answer that is given is an identification of action. It is the central idea of this book that such action identifications perform pivotal functions in a broad range of psychological and social processes.

1 An Overview

One of the fundamental human conceits is an insistence that we know what we are doing. A person may be caught in the midst of a patently ridiculous act, interrupted in a moment of apparent confusion, or even aroused from sleep, and yet respond to a query of “What are you doing?” with remarkable ease. The answer that is given is an identification of action. It is the central idea of this book that such action identifications perform pivotal functions in a broad range of psychological and social processes. Although identifications of action may not always indicate that we know what we are doing very clearly or completely, they nevertheless serve as our sole means of thinking and talking about what we do. In this sense, to identify action is to create a bridge between behavior and mind.

The nature of this bridge and the traffic that it bears are our concerns in the upcoming chapters. To provide an overview of our line of inquiry, this chapter focuses on what one can expect to learn in exploring action identification. Initially, we consider why action identification might be important, both to people in everyday life and to psychologists attempting to define and interpret human behavior. Then we will turn to our major theoretical concern—the connection between actions and their identities. We discuss the way in which the person’s understanding of action influences action, and is in turn influenced by it. This preliminary statement of action identification theory is followed by a brief survey of the kinds of questions about human behavior for which the theory suggests answers.

The Uncertain Act

Identifying an action would seem to be a simple exercise. One need only observe an act and tell what was done. Imagine, for example, that we happen to observe a friend enjoying a candlelit dinner in a small restaurant with someone we know is not his spouse. Peering from behind a potted palm, we watch him engage in an act. We turn to our companion and report the identity of the act: “He kissed her.” But our companion, an inquisitive sort, now demurs. It seems there is still some question about what was done. A conversation ensues in which it is suggested that what he was really doing was “engaging in a shameless public display.” And in just a few more moments, it is successively hypothesized that he was “offering her a greeting, ” “only touching her with his lips, ” “having a little fun, ” “making her look like a bimbo, ” “demonstrating his affection, ” and “betraying his wife of 15 years.”

Although the piano player of Casablanca reminds us that “a kiss is still a kiss, ” it appears that kissing in this case could be subject to many other identities. This is true even when we speak of a particular kiss between particular people at a particular time and place. And, kissing illustrates the rule rather than the exception. The average person with a normal vocabulary seems capable of offering multiple alternative identities for commonplace actions of every sort. “Turning on a light” can also be identified as “flipping a switch”; “going for a swim” can also be identified as “getting wet” or “getting exercise”; “attending college classes” may be seen as “working toward a degree, ” “seeking knowledge, ” or even “sitting in various chairs around campus.” Quite simply, actions do not have labels sewn in them to indicate their true identities. An identification of action is a selection of one identity from an array of potential identities, each of which might be seen in the proper circumstances as the appropriate one.

This profound uncertainty in how an action is to be identified has been recognized for some time by philosophers (e.g., Anscombe, 1957; Austin, 1961; Danto, 1963; Davidson, 1980; Goldman, 1970; Kaplan, 1964; Melden, 1961; Ryle, 1949; Schwayder, 1965; Wittgenstein, 1953), and has fueled a heated controversy among them. The question is this: If an action can be identified in many ways, and a person does it, then what did the person do? Since act identities are not just synonyms for one another, but rather depict very different meanings of the action, it is not clear whether there is one act or many. This abstract dilemma becomes more disturbing when we translate it into the question of what subjects do in psychological studies of human behavior. Subjects performed an action, for example, in the classic Milgram (1963) study of obedience. As a rule, they obeyed the instructions of an experimenter to the point of inflicting apparently dangerous shocks on another person. Or did they? Their action could also be identified, after all, as participating in an experiment, aiding in the advancement of psychology, helping someone to learn, operating a shock machine, flipping switches, or merely moving their fingers.

We do not, however, commonly refer to Milgram’s classic studies of finger movement. Nor do we talk of seminal research on line judgment (Asch, 1951), rubber doll-punching (Bandura, Ross, & Ross, 1963), or experimenter-fetching (Latané & Darley, 1970). We speak of obedience, and of conformity, aggression, and helping, and thus ignore a host of other identities that might just as rightfully be assigned to the actions that people performed in these experiments. This, then, is the problem that has bothered the philosophers—and it should be clear that it is not only a philosopher’s problem. Every research tradition purporting to explore a particular human behavior is based on the apparently arbitrary choice of one identity to represent an action that has many.

Psychologists have attempted to untangle this snarl in several ways. One option is to eliminate the concept of action from psychological theory. In this way of thinking, the things people do are identified at the most basic level— bodily movement (cf. Danto, 1963). This approach treats all of the rich meanings of action as excess baggage, and unfortunately, also denies psychologists the privilege of calling behaviors by such names as aggression or helping that are undefined at the level of bodily movement. Another approach is to stress the operational definition of action, the conceptual grounding of each abstract action identification in the concrete movements and circumstances that surround it. This approach has never gained much currency in the psychological literature because it dramatically restricts what can be called an instance of an action. At the extreme, an action can only be observed in an experiment because it is only under experimentally controlled conditions that all the detailed features that operationally define the action can be assembled.

A more common approach in recent psychological thinking is the head-in-the-sand strategy. This is a matter of ignoring the uncertainty of action and trusting that the robust scientific method will save the day. It can be argued, after all, that even though the action of subjects in a particular experiment on “aggression, ” for example, might be identified in many ways, the common denominator across many different aggression experiments will be the “aggression” identity. As science marches on, then, generalizations can be made about the nature of aggressive action. By this logic, the head-in-the-sand strategy is likely at least to be safe.

We believe these approaches entirely miss the point. What is most remarkable about the uncertainty of action in psychological research is its extreme contrast to the certainty of action that people express in everyday life. People always seem able to tell what they are doing, and it is curious that psychologists have taken so little advantage of this fact. Indeed, when an errant child attaches a suction dart to one’s forehead from several paces, one’s strongest impulse may be to cry out “What do you think you’re doing?” Psychologists have been unusually adept at suppressing this impulse, even though the actions of subjects in their experiments are often just as perplexing. Our approach to the uncertain act problem, in short, is to take it as an invitation to find out what people think they are doing.

What might be learned in asking such questions? One possibility, of course, is that we might discover the size of the gulf between what psychologists think people are doing and what people think they are doing. This is not our primary concern, however, for the point has already been well made by other commentators (e.g., Mixon, 1979). What we wish to do is to look deeper, to the wellsprings of both actions and their identities. We want to know how the person’s identifications of action may influence action. Beyond that, we want to know how the identities themselves arise. Our suspicions in this regard constitute the theory of action identification, and it is to a first general look at this theory that we now turn.

Actions and Their Identities

There are two conspicuous sources of evidence about what a person is doing. First, there is the stream of activity that is observable in the person’s bodily movements. From finger snaps and eyeblinks to dance routines and orations, there exists a seemingly continuous flow of activity in the course of the person’s life (cf. Barker, 1963). Now, within this stream of action, there is a remarkable second stream—remarkable, because it appears to be a verbal commentary on the first. This is the stream of action identification. Along with the continuous activity of the body, we find the person continuously capable of offering a report on what he or she is doing. Of course, people do not volunteer this information at every point in the stream of action, nor do they need to, for we as observers are typically schooled enough in human actions to supply our own identifications much of the time.

The stream of action identification does seem to offer evidence, however, that an observer of the stream of action might not otherwise have available. For one, a person seems able to identify his or her activities even when observers find them inscrutable and impossible to identify. The person also seems capable of identifying many actions performed in the past, and, with some success, actions that are yet to unfold. In a way, then, when we enter the stream of action identification by asking the person questions about action, we gain a portrait of much of what the person has done, is doing, and will do. This portrait is importantly connected to the stream of action itself. Or, at least so it seems. The vital question for a theory of action identification is whether the stream of action and the stream of identification are indeed interconnected.

The simple connections. A straightforward method for considering this problem is to think about the most basic kinds of connections that might exist in the streams of action and identification. Figure 1.1 displays diagrams of four simple connections that we might entertain as possibilities. In each case, identifications (IDs) and actions (As) are shown in their separate streams over time, and a dotted line is inserted to signify the relevant connection.

The identity connection shown first in Fig. 1.1 asserts no linkage at all between identities and actions. It suggests only that an identity offered at a particular time is predictive of an identity offered at a later time. If this were the only existing connection in the action and identification streams, then we could only expect that a person who says “I’m planting a shrub” at one time would be likely to say something very similar later. Unfortunately, we would have no information about any regularity in action itself, and would have to be content with studying the apparent (and seemingly meaningless) consistencies in the person’s reports of action.

FIG. 1.1. Simple connections among actions and their identities.

The action connection shown second in Fig. 1.1 again leaves the connection between actions and identities unspecified. It asserts that an action completed at one time will be predictive of an action occurring later. This is the connection appreciated by behaviorists, and often sought in studies of the consistency of behavior over time. From this perspective, the stream of identification may be assumed to be epiphenomenal, a series of yammerings emitted by the person as action ensues that has no special relationship to the occurrence of the action. Psychologists who insist on this sort of connection often are comfortable with assuming an identity connection as well—because the making of an identification is also an action (e.g., Skinner, 1957). The seeming coherence and stability of the person’s talk about action is thereby explained, but the possibility of connections between the streams of action and identification is still denied.

The intent connection (third in Fig. 1.1) is one major way in which the two streams might be linked. If this connection existed, then an identity offered at one point in time would predict the occurrence of an action later on. The identity could be volunteered well before the action occurs, in which case we might speak of the person’s planning or anticipation of the action. Or, the identity could be offered just as the action ensues; in such cases, we are less likely to ascribe “planning” to the person, but we would nevertheless credit the person with some foreknowledge of the action. It is interesting that this connection is so commonly presupposed that it is often incorporated in the very definition of action by many legal, philosophical, and psychological commentators. A person is not blamed as strongly for a faulty action, for example, if the action was not intended (Shaver, in press). This common definition of action suggests that people do not typically appreciate the possibility of the fourth connection—the one we have labeled reflective in Fig. 1.1.

The reflective connection represents the possibility that an action performed at one time may be predictive of an identification of it that is made later on. The existence of this connection suggests that a person could, for example, do a favor for a friend, and only after the onset of this action come to realize that this is what is being done. The accounts of this connection, both philosophical (e.g., Hampshire, 1959; Ryle, 1949) and psychological (e.g., Bern, 1972), paint a picture of the person as absent-mindedly unaware of action before its occurrence. Once an action begins, however, the person becomes a self-observer and attempts to find out what identity the action might have. Sociologists have remarked on such after-the-fact accounting for action (e.g., Mills, 1940; Scott & Lyman, 1968; Semin & Manstead, 1983), arguing that people exert tremendous effort to understand and justify what they have done. However, if this were the only connection among actions and identities, the stream of action would remain unexplained by reference to identities, and the stream of identification would represent only a retrospective record of what was done. This record would not be expected to differ substantially from one made by an observer watching the person without benefit of any foreknowledge of the person’s actions.

These four connections are the building blocks from which theories of action are constructed. As a rule, though, a theory emphasizes one or two of the connections and de-emphasizes others. So, as we have already seen, behaviorism highlights the habitual forms of behavior represented by the action connection, allows for the identity connection as a special case of the action connection, and downplays both the intent and reflective connections. At another extreme are the radical phenomenologists who stress the intent connection (e.g., Ashworth, 1979; Gauld & Shotter, 1977; Harré & Secord, 1972). These theorists also allow for the identity connection, arguing for a certain continuity in the person’s understanding of action (Rommetveit, 1980; von Cranach & Harré, 1982), but they standardly ignore both the action and reflective connections. A third group of theorists concern themselves primarily with the identity connection, focusing on the mental structures that would be necessary for memory of action. Structures such as “scripts” (Abelson, 1981; Schank & Abelson, 1977), for example, are proposed as ways in which people organize their memories of what they do. A fourth mode of emphasis is embraced by self-perception and attribution theorists (e.g., Bern, 1972; Heider, 1958; Jones & Davis, 1965; Kelley, 1967; Newtson, 1976). This approach to action focuses exclusively on the reflective connection, stressing the ways in which people find out about themselves and their actions. Although these theorists note that the person’s self-understanding eventually comes around to influence the person’s action (e.g., Kelley, 1980), they pay little heed to the functioning of the intent and identity connections that would be necessary for this to happen.

Our theoretical concern is with the interplay of identification and action. Quite simply, we want to know how “knowing what one is doing” can shape and be shaped by action. For this reason, our emphasis on the simple connections will also be selective. Although we will have some things to say about the operation of the action connection, we will more commonly stress those connections that involve identities in some way. But in marked contrast to the lines of inquiry undertaken by previous theorists, we will not dwell on the simple connections. We are convinced that the most useful way of understanding the interaction of the streams of action and identification is to discern how the simple connections operate in concert. Each of the simple connections, after all, stands for only one part of what appears to be a continuous, cyclic interplay of the parallel streams. We believe that the mind guides action in the flow of time, and that in this flow, each of the simple connections may be implicated in turn.

This breadth of concern means that we need not be as limited as prior theorists in our definition of “action.” Most commonly, theorists stressing one or two of the simple connections have incorporated their particular emphasis into their working definition of action. Those interested in intent say action must be defined as intentional and not accidental (e.g., Simon, 1982; von Cranach & Harré, 1982); those stressing habit say action is any bodily movement (e.g., Skinner, 1953); those focused on reflection define action as bodily movement that can be recognized (e.g., Schmidt, 1975). These definitional exercises are sometimes carried out at great length, perhaps because they are disguised polemics in favor of the theorist’s preferred simple connection (see, e.g., von Cranach, Kalbermat-ten, Indermuhle, & Gugler, 1982). But for our purposes, action can be any of these things, and so does not differ ...