June 2004336The Psychologist Vol 17 No 6Magical thinking –Reality or illusion?IS there any connection between thethought process of an adult refusing towalk under a ladder and a child scaredby a toy lion that has ‘come to life’? Howcan magical thinking protect us, orcondemn us to a life of psychologicalillness? Read on as I reveal all in an areacuriously neglected by psychology.Magical thinking and magicalbeliefsWe find it quite natural that our thoughtsor words can produce effects in our mentalworld, or in the outer world: we can thinkof moving our hand and do it, we can askfor a favour and be granted it. What wewould find surprising is if our thoughts,wishes or words produced similar effectsin non-animate physical objects. We wouldnot try to move a rock by just wishing itwould move (‘thought over matter’ magic).Non-animate objects don’t suddenlyacquire spontaneity (‘coming to lifemagic’) or violate fundamental propertiesof space and time (e.g. solid objects goingthrough each other without damage beingdone). Old people don’t become youngagain (‘transformation magic’).All this and more can happen in ourdreams and imagination. Here we are allmagical thinkers – we can be young again,fly, move ourselves to other planets in aninstant. But can we admit that magicalthinking could leave its home ground –imagination – and trespass into the worldof reality? If we do that, then we upgradeourselves from magical thinkers intobelievers in magic.In ancient Greece people believed thatthe Oracle could see the future, and thatsacrifices of animals to gods affectedweather and crops. Everybody was abeliever in magic, and the world was fullof magical transformations. Today, thingsare different. Scientists have explainedmuch of the physical world and producedeffects that would be viewed as magicala few centuries ago (transmitting auditoryand visual messages remotely; flying in theair and space). They persuaded most of usthat believing in magic contradicts botheveryday experience and the fundamentallaws of nature. In the modern civilisedworld magical thinking is ousted fromnature and finds its last refuge in art,religion and imagination. Yet traces remain,in the everyday lives of both children andadults, with some surprising and importantimplications.Do children believe in magic?Psychologists have long been fascinatedby the fact that young children in Westerncultures remain relatively free from the gripof rationality. Karl Buler (1930) wrote ofearly childhood as a period of fairy tales,when children really believe in dwarfs andgiants. Taking a more empirical approach,Jean Piaget (1926) found that four- toseven-year-old children can attributeconsciousness to non-animate things (e.gstring that ‘wants’ to untwist because ‘itknows it’s twisted’), and that desire canlead to a belief that magical activities caninfluence reality (e.g one boy believed hisailing mother would get better if he gaveup a precious toy).In a more recent study Harris et al.(1991) asked children aged four and sixyears to pretend that there was a creature(a rabbit or a monster) in an empty box.When left alone, many childrenbehaved as if the pretend creature wasreally in the box. The authorshypothesise that children ‘infuse’ thecreature with reality, believing thatthinking alone can create real physicalobjects (‘consciousness over mattermagic’).In one of my studies (Subbotsky, 1985)children aged four, five and six years weretold a story of a girl who had beenpresented with a magic box that could turnpictures into real objects. When asked ifsuch things can happen in real life, almostall children denied this. But when theexperimenter went out of the room ‘tomake a phone call’, up to 90 per cent ofchildren tried to magically convert picturesinto objects and were bitterly disappointedwhen this did not happen.In a different experiment (Subbotsky,1985) children of the same age were tolda story of a girl who had a magic table fora birthday present. The table could turn toyfigures of animals into real live ones.Again, asked if this could happen in life,only a few four-year-olds said ‘yes’. Yet,when the children had an opportunity tosee a real table that looked exactly like theone in the story and saw that a small plasticlion started moving on the table (throughthe use of magnets), only a few of thechildren behaved in a rational manner(looked for the mechanism,searched for thewires). Therest oftheDo we believe in magic? And if we do, then why?EUGENE SUBBOTSKY investigates.WEBLINKSEugene Subbotsky’s homepage:www.lancs.ac.uk/staff/subbotskMagic movements?
children either ran away fearing that thelion was coming to life, or applied a magicwand they had been given in order to stopthe lion moving. These studies clearlyshow that when children are about four,most are aware that magic can exist onlyin fairy tales. But this awareness isconfined to verbal judgements: in theiractions, four- to six-year-olds behave asmagic believers.Why do children believe in magic?There is no doubt that our culture supportsand maintains magical beliefs. Mostpreschoolers believe in Santa Claus, theEaster Bunny and the Tooth Fairy, andparents usually support these beliefs.Woolley et al. (2002) even managed toinduce some level of belief in the realityof a novel fantastic entity (the CandyWitch) in three- and four-year-old children.The fact that beliefs in fantastic entitiesare so widely spread among preschoolers,coupled with the children’s ‘magicalbehaviour’ shown in the aboveexperiments, makes it puzzling that in theirverbal judgements most children deny thatmagic can be real. Could this be due toa special balance between the costs andbenefits of engaging in magical behaviourin different circumstances, as somepsychologists believe (Woolley, 1997;Woolley & Phelps, 1994)? For instance, inan interview situation children are likely toshow rationalandlogical thinking, because an interviewerexpects this from them; whereas thinkingin a magical way brings no benefit to them.The balance of costs and benefits wasreversed in the situation when the samechildren felt in danger of being attacked bya lion if they disregarded the possibility ofthe magical transformation; and even ifthey didn’t believe, there was nobodyaround to appreciate their courage.But how stable is children’s verbalscepticism towards magic? If shown amagical effect, would children be able todefend their rational views and discount theeffect as a trick? Or would they be quick tochange their minds and say that real magiccame true? If, in this situation, theyembraced a magical explanation this wouldmean that children are ready to changetheir beliefs even when the cost–benefitbalance remains stable.To examine this, we presented five-,six- and nine-year-old children witha causal effect that looked like an instanceof real magic: an experimenter put a newpostage stamp in an apparently empty boxand cast a magic spell on the box orderingthe postage stamp to be burned (Subbotsky,2004). When the children opened the box,they found a half-burned stamp (the effectwas achieved by a trap door inside thebox). First, the children were interviewedon their understanding of the differencebetween proper magic and magic tricks.This precaution was necessary becausethe word magic can stand for both propermagic and stage magic. Before seeing the‘magical effect’, most of the children didnot believe that proper magic could happenin real life. But afterwards, most of five-and six-year-olds abandoned theirscepticism and acknowledged that thiswas an instance of proper magic, eventhough the cost–benefit balance wasthe same before and after thedemonstration. In nine-year-olds,however, only half of the childrendropped their original sceptical view,and the other half kept saying that this wasa trick. In five-year-olds magical beliefswere so strong that even after they wereshown how the trick was done, they stuckto their magical explanations. Incontrast, older children quicklyrecovered their scepticismtowards magic after the trickwas explained. The experimentconfirmed our assumption that youngerchildren’s verbal disbelief is onlysuperficial: at this age children are happyto be persuaded that magic is real. It isbetween six and nine years that childrenseem to really abandon magical beliefs.What about adults?On the basis of the above experimentsadults should be even more scepticaltowards magic than nine-year-old children.Indeed, when undergraduates wererepeatedly presented with magical effectssimilar to those described above, they didnot succumb to magical explanations(Subbotsky, 2004).Yet anthropological studies haverepeatedly shown magical beliefs to bewidespread among adults. In her account ofwitchcraft and magical practices in present-day England, Luhrman (1989) estimatesthat several thousand people, usually well-educated middle-class individuals, practisemagic. Naturalistic observations haveshown that out of 51 pedestrians passinga ladder positioned over a pavement, 37preferred to step into the road to avoidwalking under it (Jahoda, 1969). And intheir survey of magical beliefs Zusne andJones (1982) found that 64 per cent of UScollege students endorsed at least somemagical beliefs – a finding that underminesthe assumption that magical beliefs areassociated with poor science education.There is also evidence of magicalthinking in adults from psychologicalresearch. Paul Rozin and colleaguesdemonstrated that in disgust and otherdomains people’s behaviour conforms tothe main laws of sympathetic magic:contagion (‘once in contact, always incontact’) and similarity (‘the image equalsthe object’). For example Rozin et al.(1986) found that university students werereluctant to taste their preferred juice if asterilised dead cockroach was brieflydipped in it; they were also less willing totry a piece of chocolate if it was shaped inthe form of dog faeces than if it had theshape of a muffin. When given a choice,the students preferred to taste sugar waterfrom a glass labelled as ‘Sucrose’ and notfrom a glass labelled as ‘Cyanide’.Amazingly, the preference for a neutrallylabelled glass of sugar water was showneven if the alternative glass was labellednegatively (‘Not cyanide, not poison’)(Rozin et al., 1990). A similar magicaltransfer of an undesired quality froma person to an object was shown in thedomain of the fear of contagion: studentsrated a sweater briefly worn by a personwith AIDS as a significantly less desirableJune 2004337The Psychologist Vol 17 No 6Magical thinking
to wear than one worn by a healthy man(Rozin et al., 1992).The authors interpret these results asvarious cases of the same mechanism ofmagical thinking – ‘participation’(Nemeroff & Rozin, 2000). When engagedin participation, a person subconsciouslysuspends the borderline between their mind(e.g. feelings of fear or disgust) and thereal world (e.g. juice or a piece chocolatethat is perfectly suitable for consumption).Normally, participation is a usefulprotective psychological mechanism, but ifit gets out of hand it can lead to obsessivecompulsive thinking (an illusion thatexternal events which are, in fact, totallyirrelevant to a person have a personalmeaning and are intended to harm orbenefit the person).Psychological research has shown thatin healthy children there is a significantassociation between magical thinking andobsessive-compulsive thoughts andbehaviours (Bolton et al., 2002). If pushedstill further, obsessive-compulsive thinkingcan develop into obsessive compulsivedisorder (OCD). Here, magical thinkingenters the area of clinical research.Schizophrenic patients tend to engage inmagically based compulsive thinking toa considerably larger extent than both thegeneral population (Tissot & Burnard,1980) and non-schizophrenic psychiatricpatients (George & Neufeld, 1987).Altogether, these studies presentmagical thinking as scattered on a scalefrom helpful protective reactions (e.g.disgust or fear of contagion) to thereactions of a troubled mind (e.g. OCD).Potentially, studies on clinical aspects ofmagical thinking can provide insights intothe nature of hallucinatory disorders andother problems of modern life that arebased on magical mechanisms (likereligious fanaticism, ethnic conflicts orinternational terrorism). For example, thatrational people consciously do irrationalthings that bring about mass loss of humanlife, including their own, can only beunderstood in terms of magical thinking –namely, a feeling of participation in somepowerful force (God, nation, destiny) thatmakes the destructive actions seem rationalin the perpetrators’ eyes.Magical thinking is also evident insituations that involve threats to personalwelfare beyond the subconsciously basedemotional reactions in the aboveexperiments. After I cast a magic spell tobadly scratch a card in an apparently emptybox, half of the undergraduate participantsrefused to take part in a repeat with theirhands in the box (Subbotsky, 2001). Theyexplained their decisions in a manner thatsuggested they believed in the damagingpower of the magic spell. In fact, in this‘high-cost’ situation British universitystudents showed the same degree of magicalbeliefs as uneducated peasants in a mountainvillage in central Mexico, a ‘magictolerant’ culture (Subbotsky & Quinteros,2002). Yet in their verbal judgementsBritish adults – unlike Mexican adults –denied that it was possible to transformreal physical objects by a magic spell.In another experiment (Subbotsky,2003) adult participants who all denied thatproper magic could happen in real life wereasked to imagine that a woman approachedthem at dusk on an empty street. Thewoman introduced herself as a witch,and offered to cast a magic spell on hisor her future life. In one condition, thiswas a good spell, intended to make theparticipant rich and happy. In anothercondition, this was a mean spell that aimedto make the participant’s life miserable. Itwas predicted that if the participants didnot believe that their future lives could beaffected by magic, then in both conditions,for a variety of reasons, about half of theparticipants would go for the spell. But if,contrary to the scepticism shown in theinterview, participants did believe in theeffect of the magic spell on their lives, thentheir responses in the two conditions woulddiverge. In the ‘good spell’ condition, therewould still be a 50/50 split between thosewho go for a spell and those who do not.In the ‘bad spell’ condition a significantlylarger number than 50 per cent ofparticipants were expected to reject thewitch’s offer: although they would stillhave been motivated to accept the spell (inorder to comply with the witch’s request orto prove to themselves that they don’t treatthe threat seriously), the subconsciouslyheld belief that the bad spell can adverselyaffect their lives would be a force powerfulenough to outweigh the tendency tocomply. The results strongly supported the‘belief in magic’ hypothesis: in the ‘goodspell’ condition, 10 out of 17 participantssaid that they would go for a spell, eitherin order to prove they don’t believe inmagic or in order to benefit from thespell. In the ‘bad spell’ condition, all 17participants said ‘no’, and justified theiranswers by the fear that the spell mightactually affect their future lives.Altogether, the above experiments showthat in Western individuals, the belief inmagic does not disappear at the age of nineyears. Rather, magical beliefs lurk at thebottom of the mind, ready to arise at theright moment. Not only are we all magicalthinkers, we also believe in magic, at leastwith part of our minds.Why do adults believe in magic?According to Bruno Bettelheim (1977), inchildren a magical belief is fuel forimaginary role-play and fantasising thathelps children to cope with the chaos ofJune 2004338The Psychologist Vol 17 No 6Magical thinkingMany of us are tempted by the enchantment of magicMOVIESTORECOLLECTION
their subconscious desires and masterdifficult problems of life. I would add tothis that thinking and playing with magicalthings helps young children to maintain thefeeling of independence and power –something that they mostly lack in real life.But why do adults believe?First of all, because magic makes thisworld a more interesting and excitingplace. Mummified by the depressingmonotony of everyday life, many of usare tempted by the enchantment of magic.Those readers who disagree may try to findan alternative explanation for the facts thatworks of imagination like The Lord of theRings and Harry Potter becamemultimillion-dollar businesses, and thatnearly every bookshop accommodatesa spacious section of occult readings.Secondly, magic can give us a helpfulhand in circumstances that are beyondrational control. According to sometheorists, the illusion of control is a typicalfeature of the human mind and has animportant adaptive function (Langer,1975; Zusne & Jones, 1982). Although anillusion, it pushes a person towards higherachievements and helps us cope with thetroubling diversity and unpredictable natureof everyday life. Thus, when we set off fora flight, we can never be 100 per centcertain that we are going to make it. It isin this kind of situation that we resort to‘magical behaviour’, like crossing fingersor knocking on wood. In more serioussituations, like having an incurable illness,a person is even more likely to turn tomagical thinking. For those who believein God, prayer can stand for magic, but forthose who do not, the belief in magic andthe supernatural is the only way toestablish and maintain hope. Thealternative is hopelessness and despair.That is why there have always been (and,perhaps, always will be) people who claimthey have special supernatural healingpowers. In fact, contemporarypsychotherapy uses techniques that aresimilar to (or based on) those developedby magic and religion, like traditionalshamans’ techniques of autosuggestion andcreating imaginary reality for healing andother purposes (Mindell, 1993).Thirdly, magical thinking, likephenomenalistic thinking, makes the non-animate world more understandable andhumane (Subbotsky, 2000). When we arein a rush and our car won’t start, we mayspeak to it. This ‘humanising’ function ofmagical thinking is heavily exploited byadvertising: in a TV clip, a speeding carcan turn into a running jaguar, and a pieceof chocolate can take a human shape.Last but not least, magical thinkingconstitutes a foundation for the way ourindividual and social mind works (seeNemeroff & Rozin, 2000). Our emotionaland communicative reactions are literallybased on the laws of sympathetic magic.The phenomena of emotional contagion,hypnotic suggestion, magical healing, andplacebo effects are just a small sample ofthose reactions. Magical thinking isimportant for establishing and maintaininghuman relations. In love, in parenting wefrequently perform little rituals (hugging,making presents, doing small thingstogether) that, from the strictly rationalview, are unnecessary. These magicalrituals shrinking or disappearing is usuallya bad sign for the relationship.To conclude, just as rational thinkinghelps us to cope with problems in thephysical world, magical thinking comes toour aid when we deal with problems in ourpersonal, social and emotional life. That iswhy magical thinking goes well along withrational logic, and is an exciting topic forpsychological research.■ Eugene V. Subbotsky is in thePsychology Department at LancasterUniversity. E-mail:E.Subbotsky@lancaster.ac.uk.June 2004339The Psychologist Vol 17 No 6Magical thinkingReferencesBettelheim, B. (1977). The uses of enchantment:The meaningand importance of fairy tales. NewYork:Vintage Books.Bolton, D., Dearsley, P., Madronal-Luque, R. & Baron-Cohen, S.(2002). Magical thinking in childhood and adolescents:Development and relation to obsessive compulsion.British Journal of Developmental Psychology, 20, 479–494.Buler, K. (1930). Die geistige Entwicklung des Kindes. Jena:Gustav Fisher.George, L. & Neufeld, R.W. (1987). Magical ideation andschizophrenia. Journal of Consulting and Clinical Psychology,55, 778–779.Harris, P.L., Brown, E., Mariott, C.,Whittal, S. & Harmer, S.(1991). Monsters, ghosts and witches:Testing the limits ofthe fantasy–reality distinction in young children. BritishJournal of Developmental Psychology, 9, 105–123.Jahoda, G. (1969). The psychology of superstition. London:Penguin.Langer, E.J. (1975).The illusion of control. Journal of Personalityand Social Psychology, 32, 311–328.Luhrman,T.M. (1989). Persuasions of the witch’s craft: Ritualmagic and witchcraft in present-day England. Oxford:Blackwell.Mindell,A. (1993). The shaman’s body:A new shamanism fortransferring healthy relationships and community. SanFrancisco: Harper.Nemeroff, C. & Rozin, P. (2000).The making of the magicalmind:The nature and function of sympathetic magicalthinking. In K.S. Rosengren, C.N. Johnson & P.L. Harris(Eds.) Imagining the impossible: Magical, scientific andreligious thinking in children (pp.1–34). Cambridge:Cambridge University Press.Piaget, J. (1926). La representation du monde chez l’enfant. Paris:Alcan.Rozin, P., Markwith, M. & Nemeroff, C. (1992). Magicalcontagion beliefs and fear of AIDS. Journal of Applied SocialPsychology, 22, 1081–1092.Rozin, P., Markwith, M. & Ross, B. (1990).The sympatheticmagical law of similarity, nominal realism, and neglect ofnegatives in response to negative labels. PsychologicalScience, 1, 383–384.Rozin, P., Millman, L. & Nemeroff, C. (1986). Operation of lawsof sympathetic magic in disgust and other domains.Journal of Personality and Social Psychology, 50, 703–712.Subbotsky, E.V. (1985). Preschool children’s perception ofunusual phenomena. Soviet Psychology, 23(3), 91–114.Subbotsky, E.V. (1993). Foundations of the mind: Children’sunderstanding of reality. Cambridge, MA: HarvardUniversity Press.Subbotsky, E.V. (2000). Phenomenalistic perception andrational understanding in the mind of an individual: Afight for dominance. In K.S.Rosengren, C.N.Johnson &P.L.Harris (Eds.) Imagining the impossible: Magical, scientific,and religious thinking in children (pp.35–74). Cambridge:Cambridge University Press.Subbotsky, E.V. (2001). Causal explanations of events bychildren and adults: Can alternative causal modes coexistin one mind? British Journal of Developmental Psychology,19, 23–46Subbotsky, E.V. (2003).The permanence of mental objects:Testingmagical thinking on physical and imaginative realities.Manuscript submitted for publication.Subbotsky, E.V. (2004). Magical thinking in judgements ofcausation: Can anomalous phenomena affect ontologicalcausal beliefs in children and adults? British Journal ofDevelopmental Psychology, 22, 123–152.Subbotsky, E.V. & Quinteros, G. (2002). Do cultural factorsaffect causal beliefs: Rational and magical thinking inBritain and Mexico. British Journal of Psychology, 93,519–543.Tissot, R. & Burnard,Y. (1980).Aspects of cognitive activity inschizophrenia. Psychological Medicine, 10, 657–663.Woolley, J.D. (1997).Thinking about fantasy:Are childrenfundamentally different thinkers and believers fromadults? Child Development, 68, 991–1011.Woolley, J., Boerger, E.A. & Markman,A.B. (2002). A visit fromthe CandyWitch: Factors influencing children’s belief in anovel fantastic entity. Unpublished manuscript.Woolley, J.D. & Phelps, K.E. (1994).Young children’s practicalreasoning about imagination. British Journal ofDevelopmental Psychology, 12, 53–67.Zusne, L. & Jones,W.H. (1982). Anomalistic psychology.A study ofextraordinary phenomena and experience. Hillsdale, NJ:Lawrence Erlbaum.‘magical beliefs lurk at thebottom of the mind, ready toarise at the right moment’