Parenis and child specialisis are ofen concerned aboui ihe role imaginary com- panions play in childrens lives. Receni research shows ihai ihe creaiion of an imaginary companion is a common and healihy iype of preiend play. Tere are many dinereni kinds of imaginary companions, including ihose based on various iypes of props as well ihose ihai are invisible. Alihough children describe many imaginary companions as kind and obedieni, oihers ihey depici as disrupiive and unruly. When children express sirong emoiion for imaginary companions or claim io have dimculiy conirolling ihe companion, ihey may appear confused aboui ihe boundary beiween faniasy and realiiy. However, ihe auihors argue io ihe conirary: Mosi children are very clear in iheir undersianding ihai iheir imaginary compan- ions are preiend. Taylor and Moiiweiler base iheir claim parily upon sponianeous siaiemenis of children ihey inierviewed concerning ihe faniasy siaius of imaginary companions. HUmn viios unvi a unique capaciiy io love, share our lives, and even bare our souls io imaginary oihers. Imaginary oihers can iake many forms, including imaginary versions of real people, fciional characiers from books, and invenied people or animals cusiom designed io meei ihe pariicular needs of iheir cre- aior. Mosi imaginary companionscharaciers we creaie and inieraci wiih and ialk aboui on a regular basisiurn oui io be invisible. Someiimes animaiions of ioys or oiher objecis iake on such a life as well. Indeed, almosi any sori of objeci can serve as ihe incarnaiion for an imaginary oiher, as illusiraied in ihe movie Castaway, which showed a close relaiionship beiween a man who was marooned on an uninhabiied island and a volleyball named Wilson. Bui ihe capaciiy io inveni and become aiiached io imaginary oihers does noi require a lengihy hisiory of social relaiionships or exiensive experience wiih inierpersonal inieraciions. Children as young as iwo or ihree ialk io iheir siuned animals and lisien io whai ihey have io say. And when a child creaies a 2008 by ihe Board of Trusiees of ihe Universiiy of Illinois 48 A ME R I C A N J O U R N A L O F P L AY S personaliiy for a ioy or invenis an invisible friend io serve as a special friend, ihai child is engaging in a basic human urge. We choose io call ihis invenied characier an imaginary companion, a ierm we ihink more appropriaie ihan ihe imaginary playmaie someiimes referred io by oihers. While ihe creaiion of an imaginary companion consiiiuies only one of ihe many forms faniasy produciion iakes during a childs preschool years, ii may well have special signifcance. In Te Work of the Imagination, Paul L. Harris describes how imagining ihe ihoughis, aciions, and emoiions of anoiher person or creaiure provides a coniexi in which children encounier and manipulaie mul- iiple perspeciives, providing praciice for real-life simulaiions of oiher peoples poinis of view. 1 According io Tracy R. Gleason and Lisa M. Hohmann, children explore ihe possibiliiies of social inieraciion in iheir play wiih imaginary com- panions in a way ihai helps io develop iheir concepi of friendship. 2 In receni years, ihere has been a surge of inieresi in childrens imaginary companions and whai ihey mighi reveal aboui cogniiive and social developmeni. 3 In ihis ariicle, we describe some of ihe fndings of our research invesiigai- ing ihis iype of preiend play. In pariicular, we address ihe quesiion of wheiher or noi children undersiand ihai iheir imaginary companions are jusi preiend, a quesiion noi much siudied ouiside ihe work conducied in our laboraiory. Our conclusions are consisieni wiih Harriss claim ihai childrens preiend play demonsiraies iheir knowledge of realiiy, raiher ihan confusion aboui ii. Imaginary Companions Are Healthy and Common In ihe pasi, parenis, psychologisis, and oiher professionals have ofen iaken a dim view of childrens preoccupaiion wiih imaginary oihers. In pariicular, parenis unsure wheiher ihis iype of preiending is healihyhave noi always welcomed imaginary companions. Tey worry aboui whai having an imaginary companion means for a child and wonder whai role a pareni should play in guiding ihis iype of faniasy, including wheiher ihey should discourage ii. Does a childs involvemeni wiih imaginary companions suggesi ihai he or she is having dimculiy making friends: Does ii mean he or she can noi disiinguish faniasy from realiiy: Even more dire, does ii mean iheir child mighi be ai risk for psychological disorders: Te siereoiype of ihe child wiih an imaginary companion iends io be a shy, wiihdrawn child, wiih some emoiional problems, who needs io make some real friends. Tis negaiive image may owe someihing io ihe faci ihai children do indeed someiimes use imaginary companions io cope wiih problems. Afer all, ihe imaginaiion is a powerful coping resource available io ihem. Children can walk confdenily pasi a scary dog when ihere is an invisible iiger ai iheir side; ihey can ialk io an imaginary friend aboui iraumaiic evenis involving family members and know ihai iheir secreis are safe. However, preiending io have an imaginary companion occurs more ofen because ii is fun ihan because a child is in emoiional disiress. In groundbreaking research discussed in Te House of Make Believe, Doroihy G. Singer and Jerome L. Singer showed ihai children who creaie preiend friends iend io be sociable and enjoy inieraciing wiih oih- ers. 4 Research in our lab ai ihe Universiiy of Oregon and elsewhere supporis ihis view, showing ihai having an imaginary companion iends io be associaied wiih posiiive characierisiics such as advanced social undersianding, as well as being less shy and more ouigoing ihan oiher children. 3 Noi only is having an imaginary companion healihy, ii is also quiie a com- mon iype of preiend play. Exacily how common depends upon whai we call imaginary companions (are ihey only invisible friends or can ihey be special siuned animals like Chrisiopher Robins Winnie the Pooh:) and where ihe infor- maiion aboui ihem comes from (are our sources parenis, or children, or adulis who remember childhood imaginary companions:). If we consider all cases of imaginary companions creaied by children up io ihe age of seven, aboui 63 perceni of children have ihem; if we only include invisible friends and exclude ihe ones based on ioys, ihe number drops io 37 perceni. In any case, playing wiih imaginary companions remains clearly a common aciiviiy for young children. What Are Imaginary Companions Like? For almosi iwo decades we have inierviewed hundreds of children, parenis, and oiher adulis aboui ihe creaiion of imaginary companions. Our main meihod is io observe children playing in our lab, conduci separaie inierviews wiih children and parenis, compare iheir responses, and ihen resolve discrepancies by asking follow-up quesiions. One limiiaiion of our research is ihai our pariicipanis have been from primarily Euro-American, middle-class backgrounds. Neveriheless, ihe diversiiy in ihe descripiions we have collecied is enormous. Te animals and people who populaie childrens faniasy lives diner in iheir vividness, iheir personaliiy developmeni, and ihe exieni io which ihey have some basis in ihe real world. Some imaginary companions are siable and long- I magi nar y Compani ons 49 50 A ME R I C A N J O U R N A L O F P L AY S lived, and children play wiih ihem regularly, whereas oihers have a much more iransiiory exisience, drifing in and oui of a childs faniasy life. In our research, we have encouniered children whose lives were crowded wiih imaginary people and animals, none of which lingered for very long. Oiher children had only one or iwo imaginary companions ai a iime bui updaied iheir friends frequenily. Siill oiher children mainiained ihe same preiend friends for several years. Someiimes children use real world objecis as props or vehicles for imagi- nary companions. We have inierviewed children who made special friends oui of dolls, oui of a wide range of dinereni iypes of siuned animals (bears, rabbiis, frogs, dogs, monkeys, Muppeis, kangaroos, dinosaurs, hedgehogs, cows, iigers, horses, dolphins, Smurfs, Tasmanian devils, cais, donkeys, squirrels, and moose), oui of refeciions in a mirror, oui of iheir own fngers, or oui of ihe leaves on a iree. However, more commonly, imaginary companions are invisible. Some of ihese invisible friends are regular soris of girls and boys who funciion as good playmaies. Children seem io have clear menial images of whai ihese friends look like and how ihey behave. Tey have no dimculiy drawing piciures of ihem and describing iheir personaliiies. For example, ihey mighi include deiails such as ihe imaginary companion being funny, making ihem laugh, and being a good companion (e.g., we always know whai ihe oiher one is going io say). Some of ihe imaginary companions have characierisiics ihai iake ihem oui of ihe realm of whai mighi be expecied of a real child playmaie. For example, some have special capabiliiies such as being able io fy, fghi crocodiles, or perform magic. Oihers have unusual physical characierisiics, like being very small (e.g., Bainior, a very small invisible boy who is compleiely whiie and lives in ihe whiie lighi of a lamp). Alihough many are aboui ihe same age as ihe child imagining ihem, some are infanis and have io be cared for (e.g., Cream, a iiny invisible baby who lives on ihe childs hand) and some are very old (e.g., Nobby, a 160-year-old invisible businessman who visiied ihe child beiween business irips io Poriland and Seaiile, whenever ihe child wanied io ialk ihings over). Many imaginary companions are animals, frequenily wiih human charac- ierisiics such as ihe abiliiy io ialk. Some animal friends are furiher embellished wiih magical powers (e.g., a cai ihai fies) or special characierisiics (e.g., superior inielligence). For example, one fve-year-old girl described her friend Dipper as an invisible fying dolphin who lived on a siar, never slepi, and was very, very, very, very fasi. He was aboui ihe size of a regular dolphin, bui covered wiih siars and all kinds of shiny siun. Our sample includes a wide range of speciescows, dogs, iigers, iuriles, dinosaurs, mice, cais, giranes, horses, ponies, lions, elephanis, monkeys, dolphins, unicorns, bears, feas, ducks, opossums, I magi nar y Compani ons 51 panihers, unicorns, and rais. We have also inierviewed children wiih imaginary companions ihai were ghosis, angels, iwin siblings, a Cyclops, and some oiher unique creaiuresHumpiy Dumpiys moiher: an invisible ialking egg wiih spiky hair, a big round egg-like head, and a human body. When adulis ihink aboui ihe aiiraciions of having an imaginary companion, ihey iend io focus on ihe joys of having a friend who is always supporiive and helpful and consisienily loving, one who agrees wiih whai you say, does whai you wani, keeps your secreis, and provides good company. Ii seems reasonable io assume ihai a made-up friend would noi suner from ihe moodiness, siub- bornness, and oiher faws of real friends. However, descripiions of imaginary companions ofen include preiend friends who are disobedieni, bossy, argu- meniaiive, and unprediciable. Tey come and go on iheir own schedule raiher ihan according io ihe childs wishes, and ihey do noi always wani io play whai ihe child wanis io play. Tey ialk ioo loudly, do noi share, or do noi do as ihey are iold. Tey can also be a real nuisance. As one child iold us, She hiis me on ihe head and puis yoguri in my hair. Do Children Think That Their Imaginary Companions Are Real? Childrens complainis aboui iheir imaginary companions raise some fascinai- ing quesiions. For example, childrens dimculiy conirolling iheir imaginary companions, combined wiih iheir sirong emoiional aiiachmenis io ihem, could be inierpreied as suggesiing ihai children are confused aboui ihe boundary beiween faniasy and realiiy. However, research suggesis ihai children are ac- iually quiie adepi in making ihe disiinciion beiween whai is real and whai is noi. Alihough ihey ofen become emoiionally caughi up in iheir preiend play, ihis is noi unlike ihe aduli iendency io respond emoiionally io movies, books, and oiher iypes of faniasy maierial. For ihis reason, researchers such as Harris and Jacqueline D. Woolley have argued againsi ihe inierpreiaiion of emoiional responses io faniasy as evidence of faniasy/realiiy confusion. Liiile exisiing work addresses ihis issue in ihe specifc case of imaginary companions, bui in our view, children are well aware ihai iheir imaginary companions are preiend. 6 We base ihis opinion mainly on childrens sponianeous references io ihe faniasy siaius of ihe imaginary companions during our inierviews. For example, when asked where iheir friends live, children poinied io iheir heads or said in my imaginaiion. Some children emphasize iheir auihorship of ihe 52 A ME R I C A N J O U R N A L O F P L AY S characierisiics of ihe friend (e.g., I can preiend hes whaiever I wani him io be). Siaiemenis like ihese can be found in response io almosi any of ihe ques- iions in our inierview. Here are some more examples: When you wani io play wiih (friends name), how do you gei him/her io show up: Someiimes I call George in my imaginaiion and he jusi says com- ing in my imaginaiion. (fve-year-old girl) I jusi imaginaiion. (four-year-old girl) I jusi make him show up. (fve-year-old girl) She does noi ialk because shes noi a real baby. (four-year-old girl) She is really noi real, jusi a funny play bear. (fve-year-old girl) I ihink aboui her and ihen I jusi siari playing wiih her as soon as she shows up. (fve-year-old girl) How did you meei (friends name): I didni. I made ii up. (four-year-old girl) I jusi made him up in my head. (fve-year-old boy) Shes jusi my imaginary friend. (fve-year-old girl) Iis jusi preiend. (fve-year-old girl) Her is a fake animal. (four-year-old girl) Iis really jusi because iis preiend. (four-year-old girl) In one of my dreams. (fve-year-old girl) In my imaginaiion. (four-year-old girl) Hes noi in real life. (four-year-old boy) I found oui ihe way io go io Sillyland and ihais how I mei her. (four-year-old girl; ihis child laier explained ihai Sillyland is where all my preiend friends live.) Where is he/she when he/she is noi wiih you: She preiends ihai shes real by herself and wiih her parenis and wiih her broiher and wiih her pei. (four-year-old girl) He goes inio my head. (fve-year-old boy) I preiend iheyre real bui iheyre noi. (four-year-old girl) He goes in my mind and ihe world in my mind is called Neoland, I mean Pokemon Land; I have iwo lands in my mind. (fve-year-old boy) Aboui 40 perceni of our inierviews coniain such sponianeous remarks. Even when children do noi use ihe ierm preiend friend io refer io iheir com- panion, ihe preiend siaius is frequenily acknowledged up froni, as illusiraied in ihe following exchange wiih a fve-year-old girl: Aduli: Do you have a preiend friend: Child: No . . . well, I only have my house ghosi. Aduli: Your house ghosi. Is your house ghosi preiend or real: Child: Preiend. Aduli: Does your house ghosi have a name: Child: George. Aduli: And is George a siuned animal or a doll or is George compleiely preiend: Child: Hes jusi preiend. Aduli: Hes invisible: Child: Yeah. Aduli: Okay. And is George a person, an animal, or someihing else: Child: Hes jusi a house ghosi. Alihough ihese iypes of commenis are quiie common, noi every child makes ihem, so we have also carefully examined ihe iranscripis of inierviews for any indicaiions ihai ihe children were confused aboui ihe preiend siaius of iheir imaginary companions. In a siudy of eighiy-six children wiih invisible friends, one child did seem io ihink her imaginary companion was real and iwo more children were a liiile unclear (Someiimes he iurns real and he ialks real so everybody can hear him.). Tese few cases siood oui in marked conirasi io ihe oiher eighiy-ihree children (97 perceni) who showed no indicaiion of any confusion. Overall, we are siruck noi only by how much children enjoy and are engaged wiih iheir imaginary companions bui also by how frmly ihey undersiand ihai iheir friends are preiend. Conclusion Te creaiion of an imaginary companion is a healihy and common iype of pre- iend play, one pariicularly iniriguing io parenis, educaiors, and psychologisis. Tere is considerable variabiliiy in ihe iypes of friends ihai children inveni, some invisible, and some based on props such as favoriie ioys. Children inveni I magi nar y Compani ons 53 54 A ME R I C A N J O U R N A L O F P L AY S idiosyncraiic deiails and siories aboui iheir imaginary companions ihai ihey are iypically happy io share. Ai iimes, ihe conieni of ihe faniasy or ihe childs emoiional absorpiion indicaies a child may be confused aboui whai is real and whai is noi. However, children responding io quesiions aboui iheir imaginary companions ofen made explicii iheir abiliiy io disiinguish beiween faniasy and realiiy. Te same child who iells an inierviewer aboui a bossy ialkaiive elephani ihai ihe child claims io be able io see and hear, will quiie likely, in ihe nexi breaih, smile ai ihe researcher and remind her ihai ii is all jusi preiend. No1is 1. Paul L. Harris, Te Work of the Imagination (2000). 2. Tracy R. Gleason and Lisa M. Hohmann, Concepis of Real and Imaginary Friend- ships in Early Childhood, Social Development, 13 (2006), 12844. 3. Paula Bouldin, An Invesiigaiion of ihe Faniasy Predisposiiion and Faniasy Siyle of Children wiih Imaginary Companions, Te Journal of Genetic Psychology, 167 (2006), 1729; Charles Fernyhough, Kirsien Bland, Elizabeih Meins, and Max Coliheari, Imaginary Companions and Young Childrens Responses io Ambiguous Audiiory Siimuli: Implicaiions for Typical and Aiypical Developmeni, Journal of Child Psychology and Psychiatry and Allied Disciplines, 48 (2007), 10941101; Tracy R. Gleason, Social Provisions of Real and Imaginary Relaiionships in Early Childhood, Developmental Psychology, 38 (2002), 97992; Eva V. Hon, A Friend Living Inside MeTe Forms and Funciions of Imaginary Companions, Imagination, Cognition and Personality, 24 (2003), 13189; Marjorie Taylor, Imaginary Companions and the Children Who Create Tem (1999). 4. Doroihy G. Singer and Jerome L. Singer, Te House of Make-Believe: Childrens Play and the Developing Imagination (1990). 3. Marjorie Taylor and Siephanie M. Carlson, The Relaiion Beiween Individual Dinerences in Faniasy and Teory of Mind, Child Development, 68 (1997), 43633; Mariin Manoseviiz, Norman Preniice, and Frances Wilson, Individual and Family Correlaies of Imaginary Companions in Preschool Children, Developmental Psychology, 8 (1973), 7279; Alison B. Shawber and Marjorie Taylor, Invisible Friends, Personifed Objecis, and Preiend Ideniiiies: Children Who Creaie Tese Imaginary Characiers Are Less Shy Tan Oiher Children. Manuscripi under review. 6. Jacqueline D. Woolley, Tinking aboui Faniasy: Are Children Fundamenially Dinereni Tinkers and Believers from Adulis:, Child Development, 68 (1997), 991 1011.