The Illusion of Conscious Will (26 page)

How extreme can this effect become? Would people invent prior intentions for entirely preposterous actions? In another example, Moll recounts just this:

I tell a hypnotized subject that when he wakes he is to take a flowerpot from the window, wrap it in a cloth, put it on the sofa, and bow to it three times. All which he does. When he is asked for his reasons he answers, “You know, when I woke and saw the flower-pot there I thought that as it was rather cold the flowerpot had better be warmed a little, or else the plant would die. So I wrapped it in the cloth, and then I thought that as the sofa was near the fire I would put the flowerpot on it; and I bowed because I was pleased with myself for having such a bright idea.” He added that he did not consider the proceeding foolish, [because] he had told me his reasons for so acting. (153-154)

Such postaction invention of intention for hypnotically suggested actions does not always happen. Sometimes, the suggested behavior is performed but then immediately denied. Moll reported one subject who followed the suggestion to laugh but then explicitly disavowed having done it (“I did not laugh”) (145). At other times, the suggested behavior is performed and no explanation emerges. Moll gave a posthypnotic suggestion, for example, that a man use an insulting expression to him on awakening. The man awakened and then paused for a few seconds, during which his face expressed some inward struggle, and then he called out “Donkey!” Asked why he would do this, the man explained only that “I felt as if I must say ‘Donkey!’” (153). The man apparently wasn’t quite ready to invent some prior wrong for which he was reciprocating, or to assert that Moll in fact resembled a donkey, and so was marooned with no explanation.

It would be very useful to know just when it is that people feel they can or must provide such invented reports of intention or when they can just throw up their hands and say they don’t know why they did something. Part of the explanation, as Moll noted, must focus on whether there are any glimmerings in consciousness of the real (hypnotic) cause of the behavior. He mentioned that in his subjects, if the original source of the suggestion was forgotten, he was more likely to observe the invention of explanations and intentions following suggested actions. Certainly, too, there must be circumstances that cry out for intentional explanation and others that are perhaps too surprising or odd to allow a fanciful construction.

The case of posthypnotic suggestion attracted the attention of Freud and gave rise to his concept of “defensive rationalization.” Freud pointed out that each of us can, at times, be like the person who carries out a posthypnotic suggestion. Almost anyone, on being “asked why he is acting this way, instead of saying that he has no idea . . . feels compelled to invent some obviously unsatisfactory reason” (Freud, 1900, 230). The psychoanalytic theory proposes that such invention of impossible intentions occurs in everyday life chiefly when prior intentions are troubling. In this theory, the reasons people offer for their behaviors after the fact may be invented to distort or cover up the true causes, which could be left out because they are unconscious, unbearable, or perhaps just unseemly.

It is more likely, however, that people invent intentions merely because they expect them. The notion of conscious agency is so easy and handy for understanding others and oneself that it is overextended to cases when it doesn’t apply. Rather than a Freudian expression of human nature’s darker side, then, the filling in of impossible intentions may merely reflect human nature’s stupider side trying to be ideal. And, ironically, it takes human adults considerable cognitive growth and development to get smart enough to be this stupid.

Developing the Idea of Intention

We humans begin life as creatures who can’t reasonably claim to have conscious will, mainly because we can’t stop drooling, let alone make a “claim” of some kind (
fig. 5.1
). But we blossom into beings who do claim a will. How does this happen? Is it just that when we’re babies and not yet verbal, we can’t say what we intend? The developmental research on intention suggests that there is quite a bit more to it than this, in that children develop not only the ability to act but also the ability to preview or know about what it is they will do. Beyond learning to perform voluntary actions of all sorts, children must develop a quite separate range of capacities to think about those actions, reason about their actions’ causes, and experience their prior thoughts of action as intentions. Figuring out what minds are like has been called the development of theory of mind.
²
This is, of course, the central task of
mind perception,
which was discussed in
chapter 1
. A big part of mind perception for any child must include developing notions of intention, conscious will, and their interrelation (Astington 1991; Zelazo, Astington, and Olson 1999).

Figure 5.1

A nineteenth-century artist’s rendition of an automaton baby doll, which seems somehow relevant here.

To say that intention is part of a child’s developing “theory” is not to say that a child has no plans or goals prior to the theory’s development. Children do lots of voluntary actions pretty much from birth. There is evidence, for instance, that newborn infants who appear to be reaching for objects indeed are doing this, not just flailing their arms. When infants reach toward “virtual objects” seen through special glasses, and thus do not actually get to touch anything, they cry, whereas when they reach out to real objects and touch them, they don’t fuss as much (Bower, Broughton, and Moore 1970). True, their disappointing inability to talk makes it impossible for us to verify the occurrence of conscious intentions in very young infants. But it is clear that as soon as language develops, very young children (2 years old or so) say and do things that suggest the operation of mental goals, planning, and forms of self-prediction (Poulin-Dubois and Schultz 1988). They report, “Gonna go now” and then they leave, or they say “Wanna banana” and then reach for one (Brown 1973). Toddlers describe their actions in advance on occasion, and so appear to have the ability to think about their actions in advance—even in words. The moral judgment and animism findings, however, suggest that very young children are not yet up to reasoning about these ideas in the way that adults often do and that the thoughts of their actions serve primarily as signals to other people of their upcoming actions.

2.
Premack and Woodruff (1978) coined this term to refer to the tendency to impute mental states to oneself and others; their interest was in whether chimps do this. For overviews of the developmental theory of mind literature, see Astington (1993); Flavell, Green, and Flavell (1995); Perner (1991a; 1991b); Wellman (1990); and Wellman and Gelman (1992).

It seems that as children develop, their speech about their actions only gradually comes to preview what they do. The Russian psychologists Vygotsky (1934) and Luria (1961) observed this in children as the development of nonsocial speech—speech used for the guidance of one’s own action. These observers described the tendency in very young children to say what they were doing after they had done it. A two-year-old might say, “I throw bunny,” just after throwing a stuffed animal across the room. Later the child might say “throw bunny” while throwing it, and at a yet later stage, “I’ve got a good mind to throw this bunny” just before doing so. This sequence would suggest that language and thought might serve merely to describe or label action after the fact in very young children but could surface as a preview to action over the course of development. The theory suggested by Vygotsky and by Luria is that the verbal representation of the action comes to have a guiding or mediating function, helping the child think about goals and determine what to do in advance.

Although the evidence on the speech mediation theory is mixed, there is a fair amount of research indicating that conscious previews of action do occur more frequently as children get older (Zivin 1979). Thus, it would seem that the occurrence of intentions and the use of intentions to compute the experience of will might make quite a bit more sense to older than to younger children. For youngsters, after all, thoughts about action may not occur very often before their actions occur and may instead perform a kind of postmortem function. Like impulsive adults, children often do things first and think afterward.

Young children don’t seem to understand that intentions should occur
prior to
action. This was illustrated in a study by Abbott and Flavell (1996) in which children were asked to consider stories about another child. Each story had the following structure: The child in the story likes A and dislikes B; at his mother’s urging, however, he intends to and tries to attain B; by accident he actually attains A instead. For instance, he says that he likes going to a friend’s house (desired goal) and not to the skating rink, but then when mom says to go to the rink he dutifully decides to gets on a bus going there. Going to the rink, then, is apparently what he intends. However, the bus driver gets lost and, gosh, ends up at the friend’s house. The question of interest is whether this child will be judged to have
intended
going to the friend’s house.

The study dealt with very young children, and of course this creates vocabulary issues. You can’t really expect three-year-olds to understand a word like
intention.
But it does make sense that they would understand the question Where did he try to go? When three-year-olds were asked this, their usual response was to indicate that the story character tried to go, and thought he would go, to the place he preferred and to which, by a fortunate error, he actually did go. In the sample story, they said, the child tried to go to the friend’s house (where the bus driver had accidentally delivered him). By so responding, they gave no evidence of distinguishing between intention and desire. Older children made this distinction and so noted that the protagonist tried to go to the place he did not want to go (the rink). Apparently, three-year-olds set aside any sense of prior intention and accepted the lucky outcome of the action as having been just what the person “tried” to do.

This effect is not limited to perceptions of others, as children will also fill in their own intentions after action. Schult (1997) set up a situation to observe this in which children tossed a beanbag toward three colored buckets. They were asked to name in advance which color bucket they wanted to hit, and they were given a color chip to remind them of their choice. Hidden at the bottom of one bucket was a prize—a picture signaling that they had “won” (and that they could mark on a score sheet). After each time they hit a bucket, the children were asked, “Which one were you trying to hit?” The interesting case here, of course, is when the child “wins” by hitting a bucket other than the color they had said they wanted. And, indeed, when the intended bucket was missed but a picture was found, the three-year-olds answered incorrectly more often than the four and five-year-olds. The youngest children often claimed that they had been “trying to hit” the winner all along.
³

3.
It turns out that adolescents with autism also have difficulty distinguishing their own intended actions from actions they perform that just happen to create desirable outcomes (Phillips, Baron-Cohen, and Rutter 1998).

This finding fits nicely with the more general observation that young children oftenmisremembertheirimmediatelypastmentalstates. Notable examples of this can be found in studies of false belief— experiments in which children are asked to evaluate the current status of things they previously understood in a different way.
⁴
In one version of this study, Perner, Leekam, and Wimmer (1987) showed children a candy box, which on examination turned out to be disappointingly full of pencils. The child was then asked what someone else will think when they first see the box. Three-year-old children often said that the other person will think there are pencils in the box, even though this box is closed and has pictures of candy on the outside. Apparently, they do not understand that the other person’s beliefs will be false and instead somehow overextend their own current knowledge (of the pencils) in making their prediction.

You would think that children would recall that they themselves had once thought there was candy in the box and so perhaps appreciate the fact that this new child would think the same thing. But children at this age don’t even seem to grasp that their own prior mental states were different. Gopnik and Astington (1988) found that children in the false belief situation even make errors about their own immediately past beliefs. That is, after the children had been allowed to play with a candy box and learn that it contains pencils, the experimenters asked, “When you first saw the box, before we opened it, what did you think was inside it?” More than half of the three-year-olds questioned in this study said they had originally thought there were pencils in the box. This answer shows what seems to be a profound lapse in immediate memory. Whatever delight or anticipation the child may have felt in expecting candy was somehow entirely erased by the reality of the pencils.