The Illusion of Conscious Will (34 page)

It seems particularly likely that this latter factor—lack of exclusivity— is of special importance in the case of action projection during facilitated communication. After all, the whole reason for doing facilitated communication is to allow the other person to act. The stage is thus set for the interpretation of one’s own action as due to causes other than one’s own thoughts. In most cases, people probably do have thoughts about the action that quite nicely satisfy the principles of priority and consistency. They may think, “It would make sense here to type an F,” for example, and they have the thought indeed just before the F is typed. But because the assumption of nonexclusivity is so overwhelming in this setting, such causal candidate thoughts may be ignored or discounted, written off as mere perceptions of the other’s actions, for example, or understood as predictions or anticipations of what the other will do (“He ought to type an F”). The potential causal role of one’s own thoughts is thus unappreciated, and the experience of will does not surface.

This analysis suggests that the simple
belief
that the action can come from the other person is the main basis of action projection. It is the attribution to outside agency, in other words, that helps to fuel the loss of conscious will. Once the belief in outside agency is in place, the processes for interpreting one’s own action are rocked at the base. A fundamental assumption changes as the person switches modes from “my will” to “other’s will.”
²
The usual interpretive scheme, in which one looks for causes of an action in one’s own thoughts, is bypassed because the causes of the action are presumed to be the other person’s thoughts. This then may lead to a failure to interpret even the most obvious causal candidates among one’s own thoughts as intentions and instead change the focus to guessing what might be on the other person’s mind.

Part of the reason these changes take place so readily in FC must be that facilitators are strongly motivated to achieve communication. They are inclined to be sympathetic with the client, of course, and beyond this there are also pressures of circumstance that arise as soon as they agree to try to facilitate a communication. Holding the client’s hand and dearly hoping for something to come through (“Why else would I be doing this?”), one is now in a position in which a “communication” is a success and anything else is failure. Rapt attention is focused on the keyboard, and in fact facilitators often may ignore or stifle the client’s other communicative signs. Facial expressions, body language, and even moaning or crying can be entirely inconsistent with what is happening at the keyboard, and the keyboard is assumed to convey the true message (Twachtman-Cullen 1997).

Beyond this immediate motivation and situational pressure, the actual practice of facilitation has a certain shoot-from-the-hip haphazardness that further promotes action projection. In practice, facilitated communication is almost never pure or simple. Facilitators report that when they work with a client, sometimes they will guess what the client would say and try to “get them started” by typing the first letters of words. Facilitators also report that they sometimes will “finish” a word or phrase once they understand the gist of what the client is trying to say. In fact, in a fine-grained examination of a series of facilitation sessions, Twachtman-Cullen (1997) discovered that facilitation involves a startling amount of overt helping. A facilitator in one session remarked “OK, to give you a start, let’s type the word
to
” (90). Another facilitator, instructing Twachtman-Cullen on how to apply the technique, said
“You might want to lead him to the letters when you first start”
(100) and
“You should sort of go to the area of the board where he should be going”
(101). These hints at how to proceed are echoed, albeit subtly, in FC training manuals (e.g., Crossley and Remington-Gurney 1992).

2.
Or to “God’s will.” A group of ultraorthodox caregivers in Israel developed facilitated communication with autistic individuals as a means of divining communications from God. The autistic clients were housed in a group home and given daily lessons in the Talmud, and then were asked theological questions during FC. Based partly on the Talmudic tenet that God selects fools and children to be prophets (after the destruction of the Temple), the communications secured in this fashion were interpreted as having mystical significance. The technique has been highly controversial in the Jewish community (Gross 1998).

In the turmoil and awkwardness of attempted FC, then, there is much room for interpretation. Facilitation often involves a long series of effortful fits and starts, with hinting, helping, erasing and starting over, waiting, guessing, and even cajoling and scolding—interposed with occasional actions, sometimes of only a letter or two, that are felt by the facilitator to be “true” communications. The facilitator’s decision of whether a whole communication was produced by self or by the client is thus a summary judgment of literally hundreds of smaller episodes that might have been “helping” or might have been “acting” as each letter was typed. The result of this curiously syncopated interaction is an overall sense that the client must have communicated, derived largely from the initial belief that such communication would take place in combination with a consequent series of errors of interpretation of one’s own thoughts about the action.

Ventwilloquism

If belief in outside agency is the central element in action projection, it should be possible to produce action projection simply by leading people to believe that their actions are being produced by someone else. Such “throwing of the will” to another person was the focus of a series of studies by Wegner and Fuller (2000). In this research, individual college student participants were asked to attempt to read the unconscious muscle movements of another person. This situation borrowed some features of the facilitated communication setting but changed things as well. Instead of asking people to support a person’s hand and finger to spell out answers at a whole keyboard, for instance, we limited responses to two keys for yes or no answers. The participant was asked to place middle and index fingers on these keys, and then an experimental confederate serving as the “client” in this situation rested the complementary fingers of the opposite hand atop the participant’s fingers (
fig. 6.4
). This yes/no format kept us from having to interpret extended typed responses.

Figure 6.4

In the studies of Wegner and Fuller (2000) the participant placed his or her fingers on two keys marked “yes” and “no,” and the confederate/communicator rested his or her fingers on top during the questions.

Our idea was to ask the participant to try to “read the muscle movements” of the confederate as the confederate was ostensibly asked a series of questions. The confederate was introduced as a fellow college student, so no communication impairment was suggested. However, at the outset and in the presence of the participant, the confederate was admonished not to respond overtly to any questions. Rather, it was explained that both the “communicator” (our confederate) and the “facilitator” (the participant) would hear a series of yes/no questions over headphones, and that the facilitator’s job was to try to “read the communicator’s unconscious finger muscle movements” and press the key that the communicator would have pressed for each question. The facilitator was told that these movements could be very subtle and might not even be consciously perceptible, but that he or she should go ahead and respond on behalf of the communicator because “you might be tapping into something of which you are unaware.”

Now, because muscle reading is quite possible, the situation was arranged to preclude this. The questions were not even transmitted to the communicator at all. Thus, each facilitator was in the position of trying to discern answers that were ostensibly conveyed through unconscious muscle movements from someone who was, in reality, entirely blank on what was being asked. The communicator’s fingers could just as well have been hot dogs, for all the relevant muscle movement they would be making. The questions transmitted to the facilitator, fifty in all, included twenty very easy factual questions (e.g., “Is the capital of the United States Washington, D.C.?”). The first main measure of interest in our experiments, then, was the proportion of correct answers produced for these easy questions. If the proportion correct was significantly greater than that expected by chance (50 percent for yes/no questions), the facilitator was contributing to the answers. We found across six experiments that participants answered these items correctly for the confederate 87 percent of the time.

This performance wouldn’t really count as action projection if the participants judged that they were answering the questions on their own. The second key measure in these studies was of the participants’ subjective sense of how much influence (on a scale from 0 to 100) the communicator had on the answers that were generated during the session. Given that the communicator heard no questions and was moving randomly if at all, zero would be the technically correct answer. Overall, however, participants acting as facilitators attributed 37 percent of the influence for the answers to the confederate communicator. This means that, on balance, facilitators correctly attributed the action more to themselves than to the communicator. However, they did not leave the communicator out of it entirely—as was actually the case—and instead attributed over a third of the influence to this inert outside agent.
³

How did this degree of action projection occur in our simple laboratory experiments? This question should rightly be answered in two parts: First, how did people get so many answers correct, and second, how did they come to attribute this to the communicator? The first question is difficult to answer at this time because in our experiments a variety of conditions were tested to see whether this tendency to answer correctly could be changed—and very few had a significant influence. The number correct hovered around 87 percent (from 83 percent to 98 percent across studies) with all sorts of different variations, including when, in one study, we even asked participants not to touch the communicator and instead merely to “empathize” with the communicator and discern his or her answers from a distance.

3.
People who gave more correct answers to the easy questions did not necessarily tend to attribute the answers to the communicator. In fact, the correlation between the number correct and the attribution of the answers to the communicator was near zero across the studies.

It turns out that people even give a large proportion of correct answers to the easy questions
when they are told to respond randomly
. In one study, we dispensed with the communicator entirely and just asked our participants to respond yes or no in a random pattern, one answer following each question they heard. Even then, some 82 percent of answers to the questions were correct. A few participants seemed to react against the right answers, giving mostly wrong answers under this circumstance. But this number was far smaller (about 15 percent) than the majority who met the questions with correct answers even when admonished to behave spontaneously and with no discernible pattern. Apparently, there is an automatic tendency to follow questions with the right answers, and this tendency may be largely responsible for the correct answering that occurs in FC when the facilitator answers on behalf of the communicator.

One change we made in the muscle reading situation had a significant influence, enhancing this automatic correctness yet further. We noticed that in many of the instructional materials about facilitated communication, people learning to facilitate were strongly admonished not to help the client press particular keys. Instead, they were regularly instructed to provide an opposing pressure, pulling back the client’s hand at all times and thereby depending on the client for the downward motion. In the words of Schubert and Biklen (1993, 12), “When supporting the hand or arm, give constant backward pressure. Push away gently from the key-board to reduce the possibility of accidentally directing the individual to-ward specific letters.” Commenting in a related vein on one FC client, Biklen et al. (1992, 11) observed, “Mary, a twelve-year-old student who does not speak words, requires support under her hand or at the wrist to slow her down. If her pointing is not slowed, she will type seemingly unrelated series of letters.”

This theme in the FC literature suggests that the resistance or upward pressure might give rise to ironic, impulsive actions in the downward direction. As we observed in our earlier discussion of the role of resistance in automatisms, people who are trying hard not to do something may at times end up doing that very thing because of ironic processes. Unconscious search processes may target the resisted behavior and, during conditions of mental load, can actually prompt that behavior (Ansfield and Wegner 1996; Wegner 1994; Wegner, Ansfield, and Pilloff 1998). A facilitator’s tendency to produce upward pressure as part of FC could, then, provide a fertile setting for the growth of automatic down-ward strokes.