The Illusion of Conscious Will (23 page)

Pierre Janet (1889) was a great collector of such cases. He established the study of abnormal automatisms in terms of the concept of
dissociation
—the idea that the conscious executive part of the mind can become cut off from unconscious parts of the mind that produce behavior. This line of thinking about automatisms goes back far earlier among many investigators of hysteria and hypnosis and must also include Sigmund Freud and William James.
¹⁰
Janet collected and presented a variety of unusual cases as forms of dissociation—people who experienced blindness, deafness, anesthesia of various parts of the body, convulsions, possession, odd voices or sudden new habits, physical illness, and other remarkable symptoms. Intertwined with this history is the use and study of hypnosis, which has also long traded on the idea that there are special people who are particularly good at accepting suggestions to behave involuntarily (e.g., Hilgard 1965). People prone to experience dissociation were commonly believed to be especially susceptible to hypnosis, and vice versa. Taken together, these historical traditions suggest that there may be people who are prone to dissociate and thus lose conscious control of their actions.

10.
Histories of the psychiatric and psychological analysis and treatment of automatisms and related phenomena have been written by Ellenberger (1970), Gauld (1992), and Whyte (1960), and there is an excellent account of James’s interest in these topics by Eugene Taylor (1983).

If such dissociative individuals exist, they might be identified by the frequency with which they have dissociative experiences. This is the approach taken with a personality scale called the
Dissociative Experiences Scale
(Bernstein and Putnam 1986; Carlson and Putnam 1993). The scale asks questions about experiences of the loss of agency and lapses of memory and attention. For example, it asks about the experience of finding yourself in a place and not knowing how you got there, the experience of listening to someone talk and realizing that you did not hear what was said, or the experience of noticing new belongings you did not remember buying. These aren’t quite prototypical experiences of loss of conscious will, but they’re in the ballpark. There are other items on the scale that are only vaguely related to automatism, however, including experiences of hearing voices, sometimes not recognizing family members, and being able to ignore pain, so that the overall score seems to involve the general tendency to admit to odd symptoms. People who report more such experiences on the scale are inclined to be diagnosed with certain mental disorders, including dissociative identity disorder (multiple personalities), so the scale does seem to be linked with an important pathology of the will. The evidence that this is a unitary personality trait or syndrome is not clear (e.g., Sanders and Green 1994) because the various experiences are not all commonly reported by the same people. All told, this particular scale measures both more and less than what we might want to include in a test of whether people regularly fail to have the experience of conscious will.

The odd person theory fails because it is too simple. If there are dissociative individuals who fail to experience apparent mental causation, they are not all likely to do this in the same way. There are lots of avenues by which a person might bypass an experience of will. Some people might have a tendency to forget their intentions, for example, such that when they do act, they can’t grasp where the action might be coming from. Other people might have a tendency not to notice when they’ve completed actions and so fail to draw the experience of will from what they do. Yet others might experience preoccupation with certain kinds of thoughts and so draw inferences of will with great frequency to whatever behaviors they perform that are consistent with those thoughts. Another set of people might specifically ignore or suppress certain areas of thinking and thus fail to infer that any of their behaviors relevant to those areas might be willed. There may even be people who regularly fail to grasp the consistency between their thoughts and actions and thus experience the actions as less than willed.

A full analysis of the individual differences underlying automatism could yield a broad and complex range of deficits in the ability to process information about one’s own apparent causal influence. For this reason, it is important as a first step to establish the conditions under which various such deficits occur even in people with normal processing capacities. Suffice it to say at this point that the search for a general dissociationprone personality is likely instead to unearth many individual differences that can contribute to such effects. There may not be such a thing as one type of odd person, in other words, so it may be easier to attack the problem of classifying the causes of dissociation by focusing on odd situations.

Expectant Attention

One common ingredient in several automatisms is the simple expectation that the automatism will happen. This was part of Carpenter’s (1888) analysis of automatisms; he frequently used the term
expectant attention
to describe the mind-set of people who were prone to a certain automatism. The table turner expects the table to move; the dowser anticipates the rod will dip; the Ouija board speller thinks something will be learned from the board. In the cases of automatic writing, the individuals who have tried to produce this effect seem to have devoted inordinate amounts of time to sitting around, pen or planchette in hand, waiting and hoping and scribbling. A bit of expectation seems to occur with Chevreul pendulum movement, too, because even while the person holding the pendulum is trying not to move it, there may be the sense that it will move. However, ideomotor action as measured with Jastrow’s automatograph and the Jacobson EMG technique doesn’t really require any expectancy of movement, just the idea of movement. And the Bargh, Chen, and Burrows (1996) slow walking study in fact dispenses with expectant attention to the action entirely.

Expectant attention may work in several ways to produce automatisms. First, in the process of waiting to see whether a movement will occur, the usual operation of the priority principle gets undermined and obscured. When we wait for an effect of a thought, expecting that at some time the action will happen, we separate the cause and effect in time by some variable interval. This variability cannot support a causal inference, then, as there is no constant connection in time between the thought and action. You don’t get the idea that a cat caused your sneeze, for example, if one time you sneeze immediately after a cat jumps into your lap, whereas another time you sneeze only after three days of intense cat petting. By the same token, people waiting for the Ouija board to move or for the table to turn, or who are trying not to think of something as they begin automatic writing, are undermining any regularity in the timing of the relation between thought and action, and in this way are reducing the degree to which they are likely to experience the action as caused by the thought.

Another effect of expectant attention is that it leads to a kind of “trolling” for consistent action. A fishing expedition of sorts is undertaken, as actions are sought over time that are consistent with the thought. If you think about coughing for a long, long time, for example, and manage to keep attentive to this possibility, in a long enough interval you may eventually cough, clear your throat, feel a tickle in your chest, or the like, at which moment you are then prone to note the consistency between the idea and the action. However, you won’t feel that you caused the action, again because of the failure to perceive an appropriate experience of priority. The thought of coughing was far too early. Waiting for table movement or Ouija planchette movement may operate in this way. The person sits for a long while looking for something to happen in line with some idea of what could happen. In the process of looking for a particular kind of action, the person may find actions that are at least partially consistent from time to time—and so draw attention back to the fact that the action was initially thought about—but have such low levels of consistency that the sense of causation is not strong. The use of expectant attention to create automatisms creates actions that don’t happen on time just as we think of them, so we don’t feel we’ve done them on purpose.

Expectant attention does orient us toward an action that is relevant to our initial thought, however, and in this way it promotes the experience of automatism as well. It allows us to be looking in the right direction so we can know when we have done what we thought. In the several ideomotor action demonstrations that do not draw the person’s attention to the action—Jacobson’s EMG studies, Jastrow’s automatograph, or any of the priming studies by Bargh or Dijksterhuis—the person does not experience an automatism per se. The experience of conscious will was not assessed in these studies, but it is not likely that any reduction in this experience actually occurred. In these settings,the person is not attentive to the behavior at all, so the person doesn’t even note or remark on its occurrence. The behavior qualifies as an automatism merely because it happens even when the person doesn’t know it exists. With expectant attention, however, the person is specifically oriented to look for the behavior and so when it happens, the lack of conscious will is likely to be salient. It is only by comparing thought and behavior that a lapse in intention may be detected. Then, the person is likely not only to perform an automatism but also to recognize it as such.

Movement Confusion

Several automatisms depend on circumstances that obscure the nature or source of the person’s movement. If you can’t tell what you’re doing, it is hard to gauge whether you’re doing it on purpose. Movement confusion creates difficulties in the application of the consistency principle because the relation between thought and action is clouded. Comparing what we think we will do with what we have done must depend, of course, on clarity in knowing the thought, the action, and their meaningful interrelation. Consistency can be rendered ambiguous by at least three forms of movement confusion:
misdirection, amplification,
and
obscured monitoring
.

Movement confusion occurs through
misdirection
in the case of dowsing (with both Y-rods and L-rods) and also with pendulum movement. The action is redirected such that it becomes unclear how one’s initially conceptualized movement has translated into the observed automatism. With the Y-rod, for example, it seems that moving one’s wrists together or apart (a movement limited to one direction, from side to side) can yield pressure on the rod that results in rotation of the dowsing rod point up or down (a movement of rotation in a different plane). This is confusing. Similarly, pushing the pendulum in one direction becomes pulling it the opposite way. These items are simply unwieldy, and this makes it hard to tell whether one’s actions are what one initially thought they might be.

There is also an
amplification
of movement with some of these devices. Slight tilts of the L-rods become major direction swings of the pointers. Vogt and Hyman (1959) note this with the standard Y-rod: “Two ways to produce the rod’s movements are to pull the hands slightly apart or to push them slightly together. Either of these movements creates greater tension in the rod than the force of the grip. When the balance is so upset, the rod acts like a coiled spring and may straighten out with such force that the bark may literally come off in the hands of the diviner” (128-129). Gravity probably makes the rod move down more than up when it springs in this way because otherwise we would have heard of more instances of people poking an eye out while looking for water. The pendulum also amplifies slight movements, as do the slippery surfaces of the wheeled writing planchette, the Ouija planchette, and Jastrow’s automatograph. And of course, the EMG amplifiers used by Jacobson were specifically designed to amplify movement. When the person can see the result of the amplified movement, it becomes difficult to apply the consistency principle because the action seems out of proportion and inconsistent in direction with the initial idea of what to do.

Obscured monitoring
can explain the automatisms produced by circumstances that make it difficult to see what one has done. When the monitoring of action is made difficult or impossible—as with the curtained-off automatograph, for instance, Jacobson’s EMG tracings visible only to the polygraph operator, or Hull’s surreptitious thread attached to the collar— it seems that automatisms become more likely. Similarly, automatic writing may be enhanced when people look away or have their attention diverted to other tasks. It may be that when action is hard to perceive, the consistency of thought and action becomes less evident and conscious will thus diminishes. Monitoring is entirely impossible, of course, in the case of the various modern ideomotor effects we have reviewed (such as slow walking or
Trivial Pursuit
performance) because in these cases only the experimenter notices and measures the changes in behavior that occur as a result of ideomotor effects. In all likelihood, participants in these studies thus experienced no particular sense of involuntariness in their action, even though the actions arose automatically in response to the primed ideas. With no sense of what to monitor, a person is likely to experience no special feeling of automatism even though the action may well ensue from the prime.

The social magnification of automatisms occurs through all three sources of movement confusion—misdirection, amplification, and obscured monitoring. When people work the Ouija planchette together, for instance, or sit at a table to make the table turn, they may find that their slight movements combine with the other person’s movements, sometimes producing stillness but other times yielding new misdirected movements or amplifications as well. This is compounded when, as the co-actors make these minute and unconscious adjustments for each other, they don’t know just what part of the action they personally have created. With this obscured monitoring, the group performs a mystery dance,
¹¹
a collective automatism that occurs when no one has conscious and specific knowledge of what self or others are doing.