Jump to ContentJump to Main Navigation
Magic and the MindMechanisms, Functions, and Development of Magical Thinking and Behavior$

Eugene Subbotsky

Print publication date: 2010

Print ISBN-13: 9780195393873

Published to Oxford Scholarship Online: May 2010

DOI: 10.1093/acprof:oso/9780195393873.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 18 October 2019

Magical Thinking and Imagination

Magical Thinking and Imagination

(p.79) 8 Magical Thinking and Imagination
Magic and the Mind

Eugene Subbotsky

Oxford University Press

Abstract and Keywords

In Chapter 8 (“Magical Thinking and Imagination”), the issue of how magical thinking works in the domain of nonphysical (imagined) reality is analyzed. The chapter begins with an overview of studies on children's understanding the difference between physical and mental objects, and then continues with experiments on children's and adults' preparedness to accept that real or imagined objects could be changed by a magic spell. The most interesting result of these experiments was that, unlike children, adults reported that magical manipulations had little effect on both real and their imagined physical objects, but had a strong effect on their imagined fantastical objects. Further experiments showed that participants' personally significant imagined (PERSIM) objects, such as their images of their future lives, were particularly strongly affected by the experimenter's magical manipulations. This result fits within the domain of “mind-over-mind” type of magic.

Keywords:   magical thinking, magical beliefs, imagination, mental objects, permanence, fantasy and reality

The studies reviewed in the previous chapters employed only the mind-over-matter type of noninstitutionalized magical beliefs (NIMBs). In all the demonstrations of magical effects, a mental effort (a wish, a magic spell, or a magic gesture) affected a perceived physical object (a postage stamp, a driver's license, a participant's hand). But participants' imaginary, rather than perceived, objects may also be affected by magical causation. In Chapter 1, this type of magic was introduced as mind-over-mind magic. The studies that will be reviewed in this chapter explored the belief that magic can affect the objects in our imagination.

One feature that distinguishes imaginary objects (a tree that I am imagining) from perceived ones (a tree that I am seeing) is that most perceived objects are permanent. Following Kant (1929), Piaget (1954) defined the concept of object permanence as the belief that a physical object continues to exist after it disappears from the perceptual field. Piaget argued that children understand object permanence by the age of 2 years. Having this capacity enables children to represent physical objects in a mental form (as images in memory). In later years, children's imagination is amplified by the emergence of pretend play and verbal representations (Harris, 2000; Piaget, 1962).

Children's developing beliefs in object permanence have been studied in great depth by a number of researchers (for reviews, see Baillargeon, 1987; (p.80) Bower, 1971; Subbotsky, 1991b), and this has led to clarification of the concept. For instance, Johnson and Harris (1994) have proposed that children are likely to know of a number of related constraints, including constancy (inanimate objects do not spontaneously change shape or identity), permanence (inanimate objects do not spontaneously disappear or cease to exist), and noncreation (inanimate objects do not spontaneously come into existence). In most cases, these constraints relate to animate objects as well. Thus, it can be assumed that people view an object as permanent if they believe that (1) the object cannot instantly vanish without leaving any tangible traces, (2) the object cannot instantly change into a different one, and (3) the object cannot be changed by thinking or wishing it to happen without the use of physical force, tools, or actions. Importantly, the last rule protects permanent physical objects from the mind-over-matter magic (see Chapter 1).

Are Mental Objects Permanent?

A common view is that the laws of physics apply only to perceived objects, whereas imagined objects are free from physical constraints. For instance, perceived objects are supposed to be permanent and to conform to physical causality. In contrast, imagined objects are viewed as nonpermanent and capable of magical transformations. Dreams, art, and fiction provide ample evidence of imagined objects' independence from physical constraints. In dreams, objects can alter their shapes, animals can turn into humans, and inanimate objects can turn into animals (Rittenhouse, Stickgold, & Hobson, 1994). In magical worlds described in fictional accounts, people can fly and rocks can be moved by magical powers. In psychological research, imagined objects are assumed to be nonpermanent, or “inconsistent.” For instance, Wellman and Estes (1986) claim that, unlike real physical objects, mental entities “are ‘there' only when one is actively representing them” (p. 912).

Yet, even imagined objects must be subject to certain constraints: if imagined objects were totally nonpermanent, they would be impossible to conceive. Even in dreams inanimate objects do not typically change into humans (Rittenhouse et al., 1994), and in fiction magic does not always work and magical events are mixed with ordinary physical events. In myths and fairy tales, transformations of objects into other objects are constrained by the ontological proximity between these objects. For instance, Kelly and Keil (1985) demonstrated that in Ovid's Metamorphoses and Grimm's' Fairy Tales, conscious beings were much more likely to be transformed into animals than into plants or inanimate objects. The fact that imagined objects are nevertheless subject to certain constraints has been reported in past research. Undergraduate students systematically applied anthropomorphic constraints (p.81) to the imagined psychological characteristics of God (such as the ability to concentrate on only one object at a time), despite the fact that they ascribed to God omnipotent properties that would necessarily be free of such constraints (Barrett & Keil, 1996). In research with children, Woolley, Browne, and Boerger (2006) found that 6-year-olds, and to a lesser extent 5-year-olds, treated magical causal events as constrained by parameters of ordinary causality, which included priority (a cause precedes an effect) and exclusivity (an event cannot be caused by two alternative causes simultaneously). At the same time, children are not bound to apply the constraints of ordinary reality to all aspects of imagined reality. For example, 5- and 6-year-old American children (Barrett, Richert, & Driesenga, 2001) and 7-year-old Yukatek Mayan children (Knight, Sousa, Barrett, & Atran, 2004) attributed false beliefs to humans and to some nonhuman agents (such as animals and trees), yet did not attribute them to God.

Physical Versus Fictional: Domains of Imaginary Reality

When contrasting perceived reality with imagined reality, I distinguish between two ontologically different domains of imagined reality: imagined physical reality and fictional reality domains.1 Within imagined physical reality, objects and events exist that comply with the same physical and causal constraints as their perceived counterparts. For instance, if I imagine a physical item that I see in a catalog and want to buy (such as a particular painting by a favorite artist), then the imagined item has the same properties of shape, color, solidity, and permanence as its real equivalent. Even if my thinking about the item is interrupted, on resumption I am likely to think of the item as having continued to exist and not as being recreated a second time. In contrast, within fictional reality, principles of the perceived physical world can be suspended. In this world, we can dream of impossible and fantastic objects (such as a flying dog), irregular physical objects made of nonpermanent substances (an elephant made of steam, a pencil made of smoke, a person made of shadows). The common feature that unites these classes of fictional objects is that they do not have matching prototypes in the perceived world.

Research has shown that even 3-year-olds can distinguish between imagined and perceived physical objects (Estes, Wellman, & Woolley, 1989; Wellman & Estes, 1986).2 Harris, Brown, Marriot, Whittal, and Harmer (1991) reported that 4- and 6-year-olds can distinguish between perceived objects (a perceived cup), imagined physical objects (an imagined cup), and fantastic objects (a witch flying in the sky). Five-year-olds are as good as adults are at differentiating between properties of fantastic and nonfantastic characters (Sharon & Woolley, 2004). Three-year-old children understand (p.82) that, unlike perceived physical objects, imagined entities cannot be touched or seen by other people and can also be fantastic (Wellman & Estes, 1986).

Yet whether children view mental entities as permanent or nonpermanent to the same extent as perceived physical objects is unknown. Specifically, do children (and, for that matter, adults) distinguish between different types of mental entities, such as imagined physical objects and fantastic objects, in terms of object permanence? Researchers have examined similarities between imagined and perceived objects (Attneave & Pierce, 1978; Belli, Schuman, & Jackson, 1997; Ceci, Huffman, Smith, & Loftus, 1994; Freyd & Finke, 1984; Henkel & Franklin, 1998; Johnson, 1988; Johnson, Foley, Suengas, & Raye, 1988). However, in general, little is known about how imagined objects compare with perceived objects in terms of permanence or stability.

Can Magical Suggestion Change Imagination? Mind-Over-Mind Magic

As the studies reviewed in preceding chapters have shown, under certain conditions, children and adults are prepared to suspend initially strong beliefs in the permanence of perceived objects and start believing in mind-over-matter magic. The question arises of whether a similar suspension of beliefs in object permanence can affect imagined rather than perceived objects. Asking this question means replacing mind-over-matter magic with its social offshoot—mind-over-mind magic—because changing imagined objects means changing people's minds rather than altering real physical objects that exist outside of the mind. Characteristically, most objects with which existing magical practices operate are imagined and not perceived—astrology, fortune telling, palm reading, and magical healing deal with objects and events that (may) exist in the future and not in reality. Art and entertainment, too, appeal to our magical thinking by creating fantastic objects and scenarios on a mass scale. Knowledge of how magical manipulations, such as spells and rituals, affect imagined objects can help to assess the role that magical thinking plays in determining the success of these kinds of manipulative techniques.

In research, there is evidence that preschool children treat wishing as a force that can affect (change or materialize) physical objects. Some children aged 3 to 6 years believe that imagining an object in an empty container and wishing it to appear can actually create the object (Woolley & Wellman, 1993; Woolley, Phelps, Davis, & Mandel, 1999). Furthermore, children believe that their wish can affect other people's minds. In one study, 4- to 6-year-old children believed that they were able to influence another person's behavior by simply wishing the person to do something (Vikan & Clausen, 1993).

(p.83) To summarize, the question to be answered is, to what extent do imagined objects succumb to or resist magical causation compared with perceived objects? In addition to advancing our knowledge of the development of object permanence in children, exploring the permanence of perceived, imagined, and fantastic objects adds to our knowledge of children's understanding of theory of mind. Researchers have argued that children younger than 4 years have a limited capacity for understanding how the mind works. For example, young children often fail to understand false beliefs (Gopnik & Astington, 1988; Lewis & Mitchell, 1994; Perner, 1991; Wimmer & Perner, 1983), and in certain conditions, even adults make “realist errors” in judging about the mind (Mitchell, Robinson, Isaacs, & Nye, 1996). To date, research into children's understanding of the mind has been confined to exploring its cognitive operational functions, such as the capacity to act on the basis of insufficient information and to hold false beliefs. Clearly, there is more to understand about the mind. For instance, do children understand that the mind can create both accurate and distorted images of perceived objects or, indeed, completely fictional objects? Do they understand that, given differences between imagined physical objects and fictional objects, imagined physical objects are permanent whereas fictional objects are not? If children believe that mental efforts, such as a wish or a magic spell, cannot affect perceived objects, do they believe at the same time that magical causation can affect various types of imagined objects, and to what extent? In other words, along with the cognitive tasks examined within theory-of-mind research, the mind also has to undertake ontological tasks (see, for example, Boyer & Walker, 2000). Ontologically, imagined reality can be fundamentally different from perceived physical reality and contain nonpermanent fictional objects. Imagined reality can also be similar to perceived reality, in containing permanent imagined physical objects. In this sense, along with a cognitive theory of mind, children also develop an ontological theory of mind.

To a considerable extent, studying the permanence of perceived and imagined objects in older children and adults has been impeded by the absence of a suitable method. Traditionally, nonverbal tests were used for studying permanence of perceived objects in infants, with displays such as obstruction of perceived objects by other objects, invisible displacement, or replacing one object with another one behind a screen (Baillargeon, 1987; Bower, 1971; Piaget, 1937). Clearly, this method is inappropriate for older children and adults. Unlike infants, older children and adults do not view these kinds of displays as challenges to their beliefs in object permanence; rather, they view them as tricks. For example, Chandler and Lalonde (1994) showed preschoolers aged 3 to 5 years a display similar to that earlier employed in Baillargeon's (1987) study, in which one solid object was shown passing unhindered through a space occupied by another solid object. Only half of (p.84) the children called the event magical, and even these children, after further questioning, said they had meant that the event was actually a trick.

To overcome this difficulty, an alternative method was used here to test beliefs about object permanence in older children and adults. This method is a version of the invisible replacement task (Bower, 1971). It employs a trick box that causes a physical object's disappearance or transformation in such a way that these manipulations are extremely hard to explain in terms of known mechanical and other causal factors (as described in Chapter 2). Unlike stage tricks that are shown at a distance, the effects happen in participants' own hands within a simple wooden box after participants have thoroughly examined the box and acknowledged that it is empty. In addition, participants are encouraged to search in the box for the disappeared object as much as they want to. This creates the impression in participants that the effects they observed indeed violate the physical principle of object permanence.

Turning a Rabbit Into a Fish: The Mind-Over-Mind Magic Experiment

The study reviewed in this chapter (Subbotsky, 2005) addressed two related issues. First, the baseline permanence of imaginary and perceived objects was assessed. Unlike earlier studies (Subbotsky, 1997a, 2001; Subbotsky & Quinteros, 2002; Subbotsky & Trommsdorff, 1992), in which an object's concealed replacement was accompanied by the experimenter's suggestive manipulations (an effort of will or a magic spell), in the current study baseline permanence was assessed via a demonstration of the object's change without any suggestive context. Second, the extent to which mental-physical causality (magic spell or wishing) affects the permanence of different types of objects was studied. In Experiment 1, these issues were examined on imagined physical objects (an imagined piece of paper) and perceived objects (a perceived piece of paper). In Experiment 2, the same procedures were applied to test the permanence of a fantastic object (a flying dog) compared with a perceived object (a figure of a rabbit cut out of a card). In Experiment 3, the permanence of perceived objects, imagined physical objects, and fantastic objects were compared using a different type of mind-over-mind causation: instead of trying to affect an object with a magic spell, the experimenter encouraged participants to affect the object by wishing it to happen. Experiment 4 assessed the sensitivity to magical causation of imagined objects that were personally important (participants' mental images of their future lives).

Two pretest interviews accompanied the experiments. One of the interviews assessed participants' ability to distinguish between magical events and (p.85) tricks. Indeed, Woolley and colleagues (1999) reported that most 6-year-olds who judged materialization of their wishes to be magical events actually meant that they were tricks. In order to minimize the ambiguity in the interpretation of participants' answers that contained references to magic, the interview aimed to ensure that participants who gave nonpermanent answers did indeed mean a magical change of an object, and not a “trick change” involving an explicable though hidden mechanical effect such as a sleight of hand. This interview was the same as the “two wizards” interview described in Chapter 3. The second pretest interview examined whether participants could understand the difference between imagined and perceived objects. The criteria for this distinction were externality (perceived objects are “out there,” whereas imagined ones are “in the mind”), accessibility to sensations (perceived objects can be touched and seen, and imagined ones cannot), and intersubjectivity (perceived objects can be seen by other people, and objects that we imagine cannot). Wellman and Estes (1986) reported that young children are aware of these constraints.

In order to test the permanence of perceived and imagined physical objects, in the imaginary physical object trial of Experiment 1 of this study, participants (6- and 9-year-old children and university graduates and undergraduates) were shown an empty wooden box and a clean square of paper (5x9 cm), which was immediately taken from view. They were then asked to imagine putting this square of paper in the box. Next, participants were instructed to imagine opening the box and taking the square of paper out. At this moment, participants were asked to imagine that a new element—a picture of a rabbit—had appeared on the square of paper. They were then questioned with the purpose of determining whether they treated the square of paper as a permanent or nonpermanent object. The answers were scored for attributing the imagined object with permanence, from 2 (full permanence) to 0 (nonpermanence).

In the perceived object trial, the same manipulations were done with the real square of paper. Unbeknownst to the participants, in this condition the original square was replaced by another square of paper by flapping the trap door inside the box. There were also two conditions in the experiment. In the baseline condition, the aforementioned manipulations did not involve references to magic, and in the magic condition, the change in the objects was presented as a result of the experimenter's magic spell. The aim of the experiment was to find out if participants would treat imaginary physical objects as more vulnerable to magical causation than real physical objects. We predicted that (1) children would treat the objects (both imaginary and perceived) as less permanent than would adults, (2) in the magical condition participants would treat objects as less permanent than in the baseline condition, and (3) imagined objects would be treated as less permanent than perceived objects. The results are shown in Figure 8.1.


Magical Thinking and Imagination

Figure 8.1. Mean permanence scores as a function of age, trial (imaginary physical versus perceived), and condition (magic versus baseline).

The prediction was supported that 6- and 9-year-olds would be more likely than adults to suspend their belief in object permanence in a situation where a concealed replacement was made that was difficult to explain in a rational way. This result is in accord with earlier observations that the developmental phase of establishing belief in permanence of physical objects extends far beyond the age of 2 years (Subbotsky, 1991a,b).

The results also supported the prediction that the magic manipulation would move permanence scores below the baseline. Since in 6-year-olds baseline permanence scores were low, the effect of the magic spell was not evident, yet this effect was salient in 9-year-olds and adults. The data from this experiment are consistent with earlier observations that the beliefs in the efficacy of wishing on matter decline with age (Phelps & Woolley, 1994; Rosengren & Hickling, 1994; Woolley, 2000). The data are also in accord with the experiments reported in Chapter 3 (see Figure 3.4), which showed that most 5-year-olds and half of 9-year-olds drop their nonmagical explanations and return to magical explanations if shown an anomalous causal effect. Yet the data of this experiment, as well as those reviewed in Chapter 3, suggest that the decline in children's verbal magical beliefs occurs slower than it was thought, as long as many 9-year-olds, and to a smaller extent adults, were prepared to believe that a magic spell changed perceived or imagined objects.

The prediction that imagined objects would be treated as less permanent than perceived objects was not supported by the results. Instead, the results showed a considerable similarity between imagined physical and perceived (p.87) objects in terms of their degree of permanence. How might this similarity be explained?

Imagined physical objects can serve the function of representing perceived objects that temporarily move out of the perceptual field. For instance, my house is temporarily out of sight when I go to my office, but an image of my house can be used to complete a mental activity such as wondering where I might have left some important papers. This requires imagined physical objects to have a degree of permanence comparable with that of perceived objects.

One way of examining this explanation would be to replace imagined physical objects with fantastic objects. Indeed, fantastic objects do not stand for any perceived object. If the permanence of imagined physical objects can be explained by their representational function, then fantastic objects should be treated as less permanent than both imagined physical and perceived objects. Experiment 2 tested this expectation. A second aim of Experiment 2 was to examine whether a large degree of object transformation (changing three features of a perceived object—shape, size, and color) would result in larger mean scores of permanence than the small degree of transformation tested in Experiment 1 (altering just one feature—a picture appeared on a blank piece of paper). The rationale for this manipulation is based on earlier studies of phenomenalistic causal judgments in adults. Michotte (1962) reported that participants judged an object presented as a display on a screen as the same if only one of four features (shape, size, color, or spatial location) was changed. If, however, two or more features were changed simultaneously, participants believed that the object had been replaced with another. Thus, it was predicted that in Experiment 2, participants would treat imagined and perceived objects as more permanent than participants in Experiment 1 had treated the objects that varied in only one feature.

The procedure of Experiment 2 was the same as in Experiment 1, except for two differences. First, in the imaginary object trial, instead of imagining a piece of paper, participants were asked to imagine a fantastic animal (a small flying dog with wings) that transformed into another fantastic animal (a small cat with the tail of a fish).

Second, in the perceived object trial, instead of a blank piece of paper developing a picture on it, a small rabbit-shaped piece of green card stock was converted into a much larger fish shape cut out of orange card stock. This was done in order to approximately equate the degrees of change observed in both trials (one animal changed into another animal).

Results supported the expectation that fantastic objects would be treated as significantly less permanent than perceived objects (see Figure 8.2).

In all age groups, fantastic objects scored significantly lower on permanence scores than did perceived objects. Unlike in Experiment 1, in this (p.88)

Magical Thinking and Imagination

Figure 8.2. Mean permanence scores as a function of age, trial (fantastic versus perceived), and condition (magic versus baseline).

experiment there was no difference between the magic and baseline conditions. It appears that the degree of object transformation in this experiment (one animal turning into another animal) was too strong for the magic spell to be able to move the objects' permanence below the baseline. There was no difference between age groups either. This suggests that, when a degree of transformation is big, children and adults are equally likely to treat objects as permanent or nonpermanent regardless of type. The comparisons between the results of this experiment and those of Experiment 1 supported the prediction that permanence of perceived objects would increase if the degree of transformation were increased. In the baseline condition, 6-year-olds treated perceived objects that underwent a big change (size, shape, and color, this experiment) as significantly more permanent than perceived objects that underwent a small change (a blank piece of paper upon which a picture appeared, Experiment 1), and in the magic condition 6- and 9-year-olds did the same. This suggests that, generally, children were prepared to accept that a small change in a perceived object could happen or be done in a magical way (Experiment 1). Yet when the change was big (Experiment 2), children viewed it as a stronger challenge to their intuitive experience that physical objects are permanent and cannot be altered by a nonphysical force. Due to a ceiling effect, in adults, a “degree of transformation” manipulation yielded no significant increase in permanence for perceived objects.

It has been reported in earlier research that even 3-year-olds exhibit some realization that fantastic objects are less real than perceived or imagined physical objects (Harris et al., 1991; Wellman & Estes, 1986). Contrary to the a priori assumption that in this experiment fantastic objects would be treated (p.89) as significantly less permanent than imagined physical objects, children of both age groups did not show this effect; in fact, in the magic condition, 9-year-olds treated fantastic objects as more permanent than imagined physical objects. This can be explained by the difference between conceptual and ontological judgments (see Chapter 3). While being able to distinguish between concepts of fantastic and imaginary physical objects (for instance, by saying that imagined physical objects exist whereas fantastic objects do not), children may fail to realize that fantastic objects are less permanent than imagined physical objects.

Unlike children, adults did treat fantastic objects as significantly less permanent than imagined physical objects. At least two explanations of this contrast between children's and adults' views on permanence of fantastic objects are possible. First, it can be assumed that it is only in adults that the divide appears between the two domains of imagined reality: imagined physical reality and fictional reality (as discussed earlier in this chapter). Like children, adults viewed imagined physical objects as similar to their perceived counterparts; consequently, they attributed imagined physical objects with the same degree of permanence that they did perceived objects. Unlike children, adults viewed fantastic objects as a part of fictional reality, and therefore treated them as significantly less permanent than imagined physical objects.

An alternative explanation of the contrast between fantastic and imagined physical objects in adults would be to assume that adults selectively associated fantastic objects with magic (a phenomenon that most adults consider a product of fantasy), whereas children made no such association. Evidence for this explanation comes from explanations that some adults gave to the fact that one fantastic animal had turned into another one (“This is such a crazy animal that, probably, it can do such things,” “Because this flying dog does not exist anyway, it could have just changed into the cat with a fish tail”). Due to this selective association between fantastic objects and magic in adults, permanence of fantastic objects may have been undermined, whereas permanence of imagined physical objects remained unaffected by the magic spell.

In order to test the alternative explanation, in Experiment 3, children and adults were encouraged to change both types of imagined objects with the power of their own wish. For example, in the imagined physical object trial, the instruction was as follows: “Now I'd like to ask you to try and think hard that a picture of a rose flower appears on the piece of paper that you are imagining in the box.” Unlike a magic spell, which is traditionally associated with magic and fantastic things, wishing does not necessarily bear a magical connotation. Indeed, we can reject the idea that somebody else's magic spell can change physical objects that we are imagining, yet find it quite natural that our own wish can. If in Experiment 2 adults selectively associated fantastic objects with magic and on this ground treated fantastic objects as (p.90) nonpermanent, then in this experiment, in which the possibility of such an association was eliminated (participants attempted to change imagined objects by the power of their wish), adults would view fantastic objects to be as permanent as either perceived or nonfantastic imagined objects.

The results of this experiment (see Figure 8.3) did not support the idea that in Experiment 2 adults selectively associated fantastic objects with magic and on this ground treated fantastic objects as nonpermanent. In this experiment, as in Experiment 2, adults treated fantastic objects as significantly less permanent than either perceived or nonfantastic imagined objects, and children did not. Justifications of successful attempts in this experiment were similar to those given in Experiment 2: most adults said they had changed fantastic objects by the efforts of their mind, will, or imagination and qualified the change to be true magic and not a trick. This outcome supports the idea that for adults, but not for children, fantastic objects are a special type of object. For adults, fantastic objects, but not perceived or imagined physical objects, can spontaneously change or be magically converted into other objects.

To summarize, the results of this experiment support the view that, while permanence of perceived and imagined physical objects increases with age, fantastic objects remain largely nonpermanent throughout the age span explored in this study. Whereas 6- and 9-year-old children did not draw a line between fantastic and imagined physical types of objects in terms of their freedom from physical constraints, adults did. Adults believed that, whereas perceived and imagined physical objects cannot magically change into other objects, fantastic objects can.

Magical Thinking and Imagination

Figure 8.3. Mean permanence scores as a function of age and trial (imagined physical, fantastic, and perceived objects).

(p.91) The question arises as to what was special about fantastic objects that made adults treat them as more nonpermanent than either perceived objects or imagined physical objects. It may be the case that, with adults, there is a general understanding that imagined reality is divided into two ontologically different domains: the domain of imagined physical reality and the domain of fictional reality.

One feature that distinguishes fictional objects from imagined physical objects is that fictional objects have no representatives in the perceptual world. Even if illustrated in movies or objects of art, fictional objects are still sheer products of the creative imagination. To put it another way, the ontological status of fictional objects, by definition, is weaker than that of perceived objects or imagined physical objects. When two objects with a priori different ontological statuses are presented to participants, participants' reactions to these objects can show whether they can or cannot distinguish between them. Thus, a diminished ontological status of fantastic objects may explain why adults view these objects as significantly less permanent than either perceived or imagined physical objects.

As argued previously in this chapter, the domain of fictional reality includes two kinds of objects: fantastic objects (a flying dog) and irregular imagined physical objects (an elephant made of smoke). Imagined future events (a future trajectory of the flight of a butterfly) can also be included in this domain. Although a future event can become real, it is not possible to predict with certainty which of the multiple versions of this event might come true, and when it occurs, it stops being a future event. This makes an image of a future event ontologically the same as that of a fictional event.

Cursing One's Future: Magic and Personal Destiny

A special kind of imagining is a person's thoughts about his or her own future or destiny. Indeed, whatever plans we create about our future, we know that these plans can be suddenly interrupted by unforeseen circumstances. I refer to these kinds of events as “personally significant imagined” (PERSIM) objects. In addition to an individual's future life, PERSIM objects include thoughts about the future lives of close ones, the future of personally significant environments (a house, a homeland, the planet), future outcomes of risky and meaningful activities (in gambling, business, politics, war, pregnancy, marriage), and other future events closely related to an individual's health and well-being. Unlike other imagined future objects (tomorrow's weather, a TV program, or the approaching Christmas), PERSIM objects are filled with emotional significance and personal value, and this makes them ontologically weak and vulnerable to magical manipulations. Whereas fantastic objects are generally conceived in the (p.92) arts and mass entertainment, practices and persuasion techniques used in magic, religion, psychotherapy, politics, and commercial advertising target PERSIM objects. It is also an established fact that, when thinking of events or making important decisions that might affect their future lives, many otherwise rational individuals become superstitious (Jahoda, 1969; Vyse, 1997).

In order to examine whether PERSIM objects can be affected by magical causation, in Experiment 4 of this study, in the personal involvement condition, adult participants were asked to imagine that a witch had approached them in the street and said that she wanted to put a good spell on their future lives, which would make them happy and rich for life. But in order for this spell to work, they had to give their permission for the spell to be put on their future lives; without their permission, the spell would not work. The key question then followed: “If you were in this situation, would you allow the witch to put the spell on your life, or would you not? Why?” After participants answered and justified their answers, the experimenter asked them to imagine a different scenario. In this scenario, the witch was a servant to the devil and wanted to put a spell on their future lives in order to make them serve the devil as well. After this, the key question followed as before. For half of participants, the order of the questions about the good and mean spells was reversed.

In the no personal involvement condition, participants were told the same story, only this time it was happening to another person. That other person was introduced as a scientist, a rational person, and a nonbeliever in magic. Participants were then asked if the character should have said “yes” or “no” to the good and mean witches. Since in this condition someone else's PERSIM objects, and not the participants', were under the spell, participants were expected to judge the character's behavior on rational grounds only. In addition, after participants answered the questions about what the scientist should have said, they were asked what they would have said if they were in the scientist's place. This was done in order to check if making participants personally involved would make them react to the possibility of the mean spell in the same way, as did participants in the Personal involvement condition.

It was predicted that if the participants were skeptical toward magic, they would treat the magic spells as not affecting the course of their future lives—in other words, they would treat their own future lives as if they were permanent imaginary objects. For such participants, their future lives would depend on factors such as heredity and environment, and not on magic forces. As a result, these participants would be willing or not willing to allow both kinds of witches to proceed with their spells to an equal extent. Because a variety of motives can affect participants' judgments (wanting to comply with the women's request, to prove their skepticism toward magic, to avoid interfering with magic forces), these participants would be expected to give “yes” and “no” answers in about equal numbers.

(p.93) But if despite conscious skepticism toward magic, subconsciously participants did believe in the effect of the magic spell on their lives, then their responses in the two conditions would diverge. Regarding the good spell, some participants might be motivated to say “yes” in order to benefit from magic forces, or to comply with the woman's request. Other participants might be nervous about interfering with magic forces. As a result, frequencies of “yes” and “no” answers would still be split 50/50. Regarding the mean spell, while participants may not consciously believe that a magic spell can affect their future lives, subconsciously they might give this possibility some credit. This would make participants go against their rational views and make them inclined to say “no” to the mean spell with a frequency significantly above 50%. Consequently, in the personal involvement condition, the “disbelief in magic” hypothesis would predict no difference between mean and good spells, whereas the “belief in magic” hypothesis would predict a significant difference.

The results (see Figure 8.4) supported the “belief in magic” hypothesis: in the good spell condition, 10 out of 17 participants (59%) said that they would go for a spell, either to prove that they did not believe in magic or to benefit from the spell. In the bad spell condition, all 17 participants (100%) said “no,” and justified their answers by the fear that the spell might actually affect their future lives.

In the no personal involvement condition, about half of participants said that the scientist should say “yes” to both good and mean spells. The reasons

Magical Thinking and Imagination

Figure 8.4. Percent of participants who said “no” to the offer of the magic spell as a function of trial (good versus bad spell) and condition (personal versus no personal involvement).

(p.94) for allowing the mean spell were justified by pointing out that the scientist is a rational person and so should not give any credit to the mean spell (“Because if she did not believe in magic, this would not do any harm to her anyway,” “If she is a rational person, she should check it out,” “To show the witch that she did not care what she does”). Interestingly, when participants were then asked what they themselves would say to the good spell offer if they were in the scientist's place, half of them said this would be “yes” and the other half said this would be “no.” Yet, in regard to the mean spell offer, all participants who said that the scientist should have said “no” thought they also would have said “no,” and all participants who said that the scientist should have said “yes” thought that they personally would have said “no” and justified their answers with reasons suggesting that they believed that the spell could actually work on their future lives. In other words, when someone else's PERSIM objects were subjected to magical manipulations (no personal involvement condition), participants exhibited rational behavior and a disbelief in magic. Yet, when participants' own PERSIM objects were involved (personal involvement condition), participants exhibited superstitious magical behavior.

Altogether, the results of this study show the tendency of older children and adults to view perceived and imagined physical objects as increasingly permanent. This confirms the general age trend of a decrease in verbal magical beliefs in older children and adults (see works reviewed in Chapters 2 through 5). However, participants of all three age groups involved in this study viewed fantastic objects as nonpermanent. Moreover, in Experiments 2 and 3, adults treated fantastic objects as significantly less permanent than imagined physical objects, and children did not. It is hard to assume that adults simply failed to develop the view that fantastic objects are permanent, a view that they adopted regarding perceived and imagined physical objects. Rather, it is more likely that adults adopt the idea that fantastic objects are fundamentally nonpermanent. While in children the line between fictional and physical domains of imaginary reality is blurred, adults develop the view that, in the realm of fictional (but not physical) imagined reality, objects are free from physical constraints such as permanence and physical causality.

The results also suggest that magical causality has different meanings for children and adults. Children do not associate magical causality with fictional objects only; rather, they believe that magic can affect fictional, imagined physical, and perceived objects to an equal extent. For adults, magical causality becomes object specific. In regard to perceived and imagined physical objects, adults are skeptical toward the mind-over-matter and mind-over-mind magic; they nevertheless endorse the mind-over-mind magic in regard to fantastic and PERSIM objects. Looking ahead, in Chapter 11, I will be referring to the former class of objects as belonging to ordinary reality and to the latter class of objects as belonging to magical reality.

(p.95) The fact that adults endorse the mind-over-mind magic in regard to their PERSIM objects may shed additional light on why religious and magical practices persist in Western societies. Usually, such practices target PERSIM objects: people's representations about their future destiny, health, outcomes of important activities, and so forth. As long as PERSIM objects, unlike ghosts or UFOs, are undeniably real and yet vulnerable to magical causation, there will always be individuals who claim to be able to affect such objects with their magical powers. These individuals' personal intentions can vary (from an altruistic desire to “improve the world” to a more pragmatic interest of extracting financial or psychological profit), yet their methods of manipulating with people's PERSIM objects are similar. As I will argue in the next chapter, these methods are based on a special psychological mechanism—participation.

To summarize, the results of this study support the prediction of the book's main hypothesis: while, consciously, modern Western adults deny their magical beliefs, subconsciously, they believe in noninstitutionalized magic.


(1) . Boyer and Walker (2000) make a similar distinction between “the realm of imagination” and “the world of fantasy.” In this study, I refer to two kinds of imagined worlds—physical and fictional—to capture the fact that the fictional domain is based on the same mechanisms of imagination as the imagined physical domain, the only difference being that the physical constraints of the perceived world apply only to the imagined physical domain.

(2) . The term “real objects” is often used to contrast tangible physical objects with imagined objects. However, as many of imagined objects are, in some respects, real although not actually perceived, in this book I prefer to contrast imagined and perceived objects rather than imagined and real objects.