The Illusion of Conscious Will (27 page)

To explain this and other such effects, developmental theorists speak of a tendency for what is currently in the young child’s perceptual experience to outweigh other information in judgments of the mind—a kind of tyranny of present perception. The curious operating rule for very young children, then, seems to be that newly learned information is assumed always to have been known. In other words, although they may well be cute, children are lousy historians. Even when four and five-year-olds are taught new facts about animals, for example, or are taught new names for colors, they often later insist that they knew these things from the outset (Taylor, Esbensen, and Bennett 1994). They make little distinction between the novel and the familiar and so become prey to a “knew it all along” effect.
⁵

4.
The first such study was by Wimmer and Perner (1983). Reviews of this literature can be found in Gopnik (1993), Perner (1991b), and Wellman (1990).

By this principle, any presently perceived situation permeates back-ward into estimates of the past mind. It makes sense, then, that young children make mistakes about their own mental previews with great regularity. Their view of what they had intended to do in a particular in-stance is based profoundly on what they found out they did. It is only with the eventual development of a representation of one’s own mind, and how that mind might differ from one time to the next (Ingvar 1985), that the individual becomes capable of understanding that prior intentions might differ from currently understood consequences of action. The fact that adults can sort out prior intentions from the observed effects of their actions means that they must have constructed a representation of their own minds that has separate pockets for these things—a mind past and a mind present. Very young children seem to have just one big pocket.

Human adults, of course, start out as kids. This means that every one of us begins our lives with this odd tendency to invent false intentions after the fact. It wouldn’t be surprising if we slipped back into this habit from time to time or perhaps even did it quite regularly when we knew no other grownups were looking. Still, we know it is incorrect. The development of the idea of an agent gives adults a perspective with which we can see that intentions must have at least one special characteristic: They can’t occur for the first time after the action is over. To adhere to our notion of the ideal human agent, we presume that conscious intentions must occur prior to the action they apparently cause.

Unconscious Action

Not all human actions begin with conscious intentions. Much of what we do seems to surface from unconscious causes, and such causation provides a major challenge to our ideal of conscious agency. The unconscious causation of action is the fly in the ideal agent’s ointment, the blemish on what would otherwise be a well-formed conscious goal seeker. So, even after children develop the idea of intention and apply it effectively to much of what they do, they encounter regular instances in which they don’t seem to have intended their action. These instances prompt a fix-up routine, an attempt to return to the ideal. To get a sense of how this happens, it is useful to consider unconscious action itself in detail.

5.
Adults do this a bit as well but not nearly as much (Begg et al. 1996).

How would you know if you were doing something unconsciously? If you do something consciously, of course, knowing seems to come with doing. But if you do something unconsciously, you are by definition out of the loop, perhaps only to become aware of what you are doing after the fact. You could be doing a bunch of different things unconsciously this minute, perhaps to find out later, or more likely, never to know at all. Right now, are you ignoring that it’s time to phone someone you promised to call? Are you sitting with poor posture? Damaging your eye-sight by reading in low light? Making the person you borrowed this book from mad that it hasn’t been returned? Twiddling with your hair or clothes in an absentminded way? Failing to get the exercise you promised yourself to get today? Obviously, you can’t be conscious of everything you’re doing—even holding perfectly still can be a variety of acts, not all of which you may know or understand. It is inevitable that you are doing many things unconsciously.

With this in mind, the odd example of a person responding to posthypnotic suggestion may not seem so odd at all. Each of us does many things throughout our lives that we might as well have been hypnotized into doing because we will never know we did them. This ghost army of unconscious actions provides a serious challenge to the notion of an ideal human agent. The greatest contradictions to our ideal of conscious agency occur when we find ourselves behaving with no conscious thought of what we are doing. When life creates all the inevitable situations in which we find ourselves acting without appropriate prior conscious thoughts, we must protect the illusion of conscious will by trying to make sense of our action. We invent relevant thoughts according to the template that conscious agency suggests. We may rue that we did not have the appropriate thought beforehand, or even without evidence that the thought was absent, we may simply assume that it was fully in place all along.

Unconscious action is possible because actions and our knowledge of them are different things. Action and knowledge of action stand in the relation of object and representation. This means that while actually hugging a puppy is an action, anything at all that reminds us of hugging a puppy which is not itself hugging a puppy can qualify as a representation of the action.
⁶
This separability of action and its mental representation means that acts can be done that we never knew (because we never encountered or thought about what they were); acts can be done that we weren’t thinking about (because we didn’t have our existing knowledge of them in consciousness); and acts can be done that we can’t think about (because we can’t remember what we may have intended or meant by them). Unconscious action results from problems, then, in
action identification
, in
thoughts of action,
and in
memory for intention
. It is important to examine each of these processes to see just how it is that unconscious actions can arise. These are what our ideal of conscious agency must be protected from.

6.
There are two immediate complications here. The first complication is that actions can themselves be representations. Bruner (1964) called this
enactive representation
and pointed out that sometimes when we don’t have words or images for something we may act it out, perhaps not even for communicative purposes. So there could be cases when doing something serves as a representation of doing something else. This need not worry us much because the act that serves to represent another act serves most of the functions of a thought and is usually quite distinguishable from it. When I grimace as a way of thinking about lifting something heavy, for example, the grimacing is easy to distinguish temporally and functionally from the lifting.

The second complication is that thoughts often have the features of actions. When you purposefully study a poem to memorize it, for example, you perform a series of mental actions—looking at the poem repeatedly, saying it to yourself, putting it down and trying to see whether you can say it again, and so on. This thinking and rethinking certainly seems to qualify as voluntary action by several criteria. Thus, it is possible to suggest that a thought about thought could sometimes look just like a thought about action. In the case of the thought about thought, of course, the first is a representation of the second. Having the desire to think about going to the beach, for example, involves a representational thought and an “active” thought.

All this is to say that the distinction we need to make here between thought and action is better understood as a distinction between either physical or mental representation and either physical or mental behavior—but thought and action are nicer words.

Action Identification

Most human actions are open to multiple identifications or descriptions. An action one might call “sunning on the beach” could perhaps also be described as “vacationing” or “getting a tan” or “lying down” or “relaxing” or even “exposing oneself to harmful cancercausing ultraviolet rays”; the list continues indefinitely. It is not just that different identifications of actions are synonyms because each identification can suggest a unique meaning of the action. “Walking into a party,” for instance, is not a synonym for “winning a door prize,” even though both may identify the same action under the right circumstances. This indeterminacy of action has interested philosophers for some time (Anscombe 1957; Goldman 1970), and it suggests a variety of psychological observations that Robin Vallacher and I have studied under the rubric of
action identificationtheory
(Vallacher and Wegner 1985; 1987;Wegnerand Vallacher 1986).

The central idea of this theory is that whereas people may think about any action in many ways, they typically think about an action in just one way. Although the person could be said to know the action through all its various descriptions, the theory proposes that the person’s effective knowledge of the action at any one moment is limited to one identity— usually the identity that the person has in consciousness or has most recently held in consciousness. The person’s conscious identification of the action can range, then, along a dimension from low-level identifications that indicate how the action is done (“I’m waving my hand”) to higher-level, more encompassing identifications that indicate why or with what effect the action is done (“I’m signaling the waiter to bring on the cheese dip”). This flexibility in the naming of actions suggests that they might be undertaken under one identity and later recognized under others.

Consider the case of the action of “shooting a person.” A burglar might go to an empty home with the conscious plan of stealing a TV and be carrying a gun in case he might need to protect himself. Hearing noises in the next room, he pulls out the gun. At this moment, if the homeowner steps into the room, the burglar might think of the next action in many ways. He might “protect himself,” “aim at the sound,” “squeeze the trigger,” “commit a felony,” “take a human life,” “shoot someone,” “make a mess,” “keep from getting caught,” or yet more things—all in the same action. Now, it is widely believed, and rightly so, that “
to the extent that someone is paying attention to their behavior,
they do not normally allow themselves to perform actions without reason” (Marcel 1988, 146-147). Yet the burglar in this case might merely be attending to the behavior of “aiming” or “squeezing” and so pretty much miss the point of what the action is all about. The action could well be murder, and in this sense, it is committed without reason.

A moral observer might be inclined at this moment in the burglar’s life to shout out, “For God’s sake, you idiot, you’re about to kill someone.” But this particular facet of the action’s meaning might be quite in the background for a nervous burglar in a dark room fumbling with the trigger of an unwieldy pistol. Thus, an action that started out in the early planning stages as “protecting myself” might be achieved in the heat of the moment merely as “pulling the trigger” and understood only at some point later in terms of its larger meaning and moral overtones as “taking a life.” Although the burglar may have done all the things that would allow his behavior subsequently to be viewed as intended, the explicit intention corresponding to the most important identification of his action—at least to someone reading the lurid story in the paper later on— may not have been in his mind beforehand or as the action was done.

In a number of experiments, it has been found that people do change their identifications of actions over time in just this way. Wegner, Vallacher, and Kelly (1983) found, for example, that people who are getting married will commonly identify what they are doing—if asked far in advance of the ceremony—in a romantic sense such as “showing my love.” Yet if they are asked to describe the act of getting married a day or two before the wedding, love is no longer in mind and instead they mention all the details (“getting flowers,” “finding the proper outfit,” “walking down the aisle,” etc.). And if they are queried some weeks after the wedding, they see it differently again, usually as “getting in-laws” or “becoming a member of a family.” The identification of action seems to change over time, often following this very pattern—from a meaningful identification, through a morass of details, and back to a meaningful identification that may or may not be the same as the original one.
⁷
People in advance of getting married might not want to think that they are merely getting in-laws, and those who are sadder and wiser afterwards might no longer adhere to their initial romantic idea of the act as showing love. However, the transition from one meaning to the other can occur through a kind of deconstruction of the act’s meaning that happens on becoming immersed in all the thoughts of how (rather than why) it is done.

7.
This process of identification change is illustrated in several studies, for example, Wegner et al. (1984).

The job of discerning intentions, in this light, can be excruciatingly difficult. The particular identity of an act can shimmer and fluctuate over time, touching on different meanings at a rapid rate.
⁸
Intentions may only hold still to be named and catalogued in those special moments when we are called upon to produce publicly reportable meanings for action. This might happen when we are specifically asked to explain why we did something or when we self-consciously pause to introspect on the meaning of what we have done. Or intentions might come to mind at the right point just before action and so fuel the inference that they caused the action as it was identified. In essence, the fact that there are different identities for any action suggests that people will inevitably be doing some things unconsciously no matter what. Unless they act very, very slowly and think about it so much their heads hurt, people are doomed to do many things they don’t consciously consider, and they will do this every single time they act.
⁹