Much of the early research on children's pretend play focused on the young child's ability to generate pretend actions. This research tradition owes its origins to Piaget's (1945/1962) seminal observations of his own children at play. Rich in description, and accompanied by a strong conceptual framework, Piaget's work inspired a series of subsequent studies of children's ability to produce pretend actions during the course of solitary play (for reviews, see Fein, 1981; Bretherton, 1984). Frequently, these studies drew on Piaget's notion of the emerging semiotic function, in particular the young child's ability to distinguish between signifier (e.g., wooden block) and signified (e.g., bar of soap), to explain the accomplishments witnessed in solitary pretend play (Nicolich, 1977; McCune-Nicolich, 1981).

While not repudiating Piaget's approach directly, a different tradition emerged in the latter part of the twentieth century. The focus of researchers in this tradition is pretend play in a social context, particularly pretending in mother-child and child-child dyads. The accompanying conceptual framework, though less explicit than Piaget's, emphasizes the social-cognitive gains that derive from collaborative interactions (see Rogoff, 1990; Cole, 1996). To be sure, an emphasis on social context in symbolic (pretend) play is in itself not new, owing a debt to Vygotsky (1930–1966/1978) and Werner & Kaplan (1963) among others. However, more recent empirical work on pretending in context has added important information about the particular factors that influence social pretend play.

Awareness of non-realities

Our capacity for fantasy has often been touted as one of the unique attributes that defines us as human. As early as 18 months many human children have entered the realm of fantasy through their engagement in pretend play. Although we are not certain of this, even younger infants may enter this realm via their dreams. And certainly way into old age we enjoy reading novels, going to plays, and engaging in elaborate day-dreams. The question addressed in this chapter is the nature of our conscious understanding of this fantastical realm, in particular the understanding that young preschool-age children display. The understanding of this age group is the focus, because many believe that at this age children begin to consider the fantastical and to make a conscious division of their world into real and not-real, a distinction that is basic to and deeply entrenched in our adult thinking.

There are many levels of reality and many varieties of non-realities. As Austin (1962, p. 67) put it, “That may not be real duck because it is a decoy, or a toy duck, or a species of goose closely resembling a duck, or because I am having a hallucination.” In previous work (Woolley & Wellman, 1990) I have investigated the extent to which children talk about the many levels of reality in their everyday conversation.

Introduction

Eye-covering play, deliberately closing or covering one's eyes during a play sequence, has been documented in various nonhuman primates. Some observers have suggested that eye-covering play involves pretending – acting as if one can't see, doesn't exist, or exists in some altered form (Cunningham, 1921; Hahn, 1982; de Waal, 1986b, 1989). Only one systematic study has been made of eye-covering play in nonhuman primates (Thierry, 1984) and no systematic attempts have been made to investigate the cognitive processes involved. We systematically studied eye-covering play in captive orangutans (Pongo pygmaeus) and Japanese macaques (Macaca fuscata) with the aim of assessing its cognitive implications in relation to pretending and imagination.

Cunningham (1921) was perhaps the first to describe eye-covering play in a nonhuman primate – in an immature lowland gorilla (Gorilla gorilla) who would shut his eyes tightly then run around knocking into furniture, which was interpreted as pretending to be blind. Eye-covering play has since been reported in all the great apes. Their play often takes the form of stumbling or groping about while “blind” (Hoyt, 1941; Harrisson, 1961, 1962; Lang, 1963; Gautier-Hion, 1971b; Goodall, 1971; Rensch, 1972; de Waal, 1986b, 1989). Occasionally they use objects to cover their eyes (Harrisson, 1962; Lang, 1963; Goodall, 1971; Rensch, 1972; de Waal, 1986b, 1989) and eye-covering play sometimes occurs as a social game (Hoyt, 1941; Harrisson, 1962).

Alice's world through the looking glass, though actually a world of dream, is in some respects like the world of pretend. Just as Humpty-Dumpty can be master of words, making them mean what he wants them to mean, in pretend play children designate what various objects and activities denote. Pretending is a special frame that organizes the activities within it. What goes on in the frame has its own reality, stipulated as the players choose, such that a behavior acted while pretending may have alternative significance while not pretending (Bateson, 1955/1972). This impressive ability to deal in framed worlds is fundamental to pretense (Fein, 1981; Bretherton, 1984). Framing also appears fundamental to other human activities, such as reflecting on dreams and understanding others' minds. Like pretense, another's mental world is framed by that person's prespective. When children learn to pretend, they learn to deal with worlds that are framed differently from the real world. In the looking glass world, nursery rhyme characters come to life, words mean anything you want, and memory can work both backwards and forwards.

The pretend world is not entirely unconstrained.

Pretend play is a fascinatingly complex behavior from which psychologists have drawn information on a wide range of children's functionings and developments. Piaget considered it as an important window through which to glimpse the incipient representational capacities of the child. Indeed, in pretend play, the abridged and schematized enactment of activities (or events) outside of their habitual context consists in the signifier, the trace evoking the real activities, the signified. The playful attitude, the abridged actions, enacted sometimes in an exaggerated manner, the inanimate co-participants, the miniature objects, the repetition of actions lacking material results, the simulation of physical sensations in the absence of relevant physical stimuli, the displacement of the activity relative to its habitual setting, etc., are all indices that have been taken, separately or in combination, by authors trying to define and/or identify early manifestations of pretend in very young children (e.g., Piaget, 1945/1962; Inhelder et al., 1972; Nicolich, 1977; McCune-Nicolich, 1981; Veneziano, 1981; Musatti, 1986; Lillard, 1993a).

More recently, pretend play has attracted researchers' attention for the potential developmental links it may have with components of “theory of mind.” Indeed, the representational aspect of pretend play implies children's ability to consider one object as simultaneously having the properties it has in “real” life and those that it has by virtue of the meaning transformation it has undergone in pretend. The ability to hold double representations about a single entity is also necessary for a theory of mind.

What could be more obvious as examples of pretense? In the first two examples, the pretender playfully attributes animate qualities to a blanket and toys and, in the second three, the pretender treats nothing as a toy, as tea, and as food. If Viki, Chantek, Michael, Koko, and Austen were human toddlers, most observers would not hesitate to call these episodes examples of “pretend play.” But these are not human children, they are great apes: chimpanzees (Hayes, 1951; Savage-Rumbaugh, Shanker & Taylor, 1998); an orangutan (Miles, 1990); and lowland gorillas (Patterson & Kennedy, 1987). In this chapter, we present evidence of pretend play by great apes, focusing on one in particular – the sign-language-using gorilla Koko.

The collection of articles herein focuses on the mysterious liminal region that lies between pre-symbolic and symbolic abilities. The transition between the two remains the least charted area, most intriguing of all developmental transformation in human childhood, and the least understood of all transformations in hominoid evolution. The mystery is deepened by disagreements over terminology. Simulation, imitation, pretense, symbolic play, representation, meta-representation, theory of mind, intentionality, and imagination, the very definition of these terms is contested territory. Authors of articles in this volume do not simplify the task because many of them disagree on these matters. Rather, their articles provide readers with a fascinating array of perspectives on these and related concepts. They also provide comparative data on a rich array of great ape species: humans, bonobos, chimpanzees, gorillas, and orangutans, plus some macaque monkeys.

Following in the wake of his earlier work on deception, self-awareness, and anthropomorphism, Robert Mitchell's new collection carries us further into contested twilight zones between infancy and childhood, and between other great ape and human minds. The juxtaposition between animals and children in the title is more than accidental since many of the same frameworks have been used to study and compare children of our species with those of our closest living relatives, the nonhuman great apes, an approach that has come to be known as comparative developmental evolutionary psychology (Parker, 1990).

Monkeys live in a complex social nexus and rich foraging habitat, and their social and intellectual skills develop in these arenas. Their tooluse and problem solving abilities put them on a different level from most mammalian clades (Beck, 1980; Antinucci & Visalberghi, 1986; Tomasello & Call, 1997; Parker & McKinney, 1999). Young monkeys must learn how to obtain sustenance, as well as how, when, and with whom they should groom, communicate, and play. They must learn to understand facial and vocal communication, recognize and manage their social bonds of kinship and friendship (even if the social ranks are asymmetric), and manoeuver in the field of their dominance relationships (which are asymmetric except for the highest and lowest ranking) (van Hooff, 1967; Stein, 1981; Fedigan, 1982; Gouzoules, Gouzoules & Marler, 1984; Smuts, 1985; Strum, 1987; Whiten & Byrne, 1988; Cheney & Seyfarth, 1990; King, 1994; Zeller, 1994, 1996).

Given the complexity of their world, the psychological complexity of monkeys is to be expected. When I watch free-ranging monkeys using directional gaze to signal an interest in food or approaching danger, or blocking others' views of a yawn (which could be perceived as a threat), I get the strongest impression that they comprehend the effects of their own actions and act intentionally. Take the following observation:

There are a bewildering number of opinions about what constitutes pretend play and the extent that nonhuman animals are capable of it. Play fighting and object play in mammals and birds strike many observers as having the “acting as if” quality that we associate with pretending. Other researchers are more inclusive, claiming that “in all play, both animal and human, there is an element of ‘pretending’ ” (Aldis, 1975, p. 14). According to Darwin (1871/1896, p. 69; our italics), “even insects play together, as has been described by that excellent observer, P. Huber, who saw ants chasing and pretending to bite each other, like so many puppies.”

When acts of pretense are assumed to reflect a metacognitive understanding of behavior, claims about pretending in nonhuman animals (e.g., Bateson, 1955/1972) become more controversial (Mitchell, 1991a). For example, Rosenberg (1990) has argued that third-order reasoning about intentionality (i.e., animal A plays with the intention of animal B understanding that A's intention is nonserious) is beyond the intellectual capacity of any nonhuman species. However, according to Bekoff & Allen (1998) such demanding criteria might be said to preclude the possibility of pretending in young children as well! In fact, a growing number of developmentalists are claiming that too much cognitive credit had been attributed to child pretenders. In particular, Lillard (1993b; 1994; 1996; PIAC7) has shown that young children have difficulty when asked questions requiring them to reason about the representational aspects of pretending.

Interest in animal behavior suggestive of pretense and imagination followed directly from Darwinian ideas about the evolutionary relationship between humans and other animals. This chapter presents a history of ideas about, and evidence of, animal and childhood pretense. After describing the scientific climate about animal pretense prior to Groos' (1898) The play of animals, I present Groos' ideas and the scientific response. I then articulate ideas about pretense from three approaches: developmental psychology; its evolutionary offshoot in ape–human comparison; and ethology.

Evolutionary implications

Given the desire which Darwin's (1859/1902; 1871/1896) theory of natural selection produced by the late 1800s to see humanity in animals' activities, it is not surprising that, from the fact of simulation, a knowledgeable “pretense” was at times inferred. Both animals' and children's play exhibit “simulated actions in place of real actions” (Spencer, 1878, p. 630), a “dramatizing” of unsatisfied instincts or adult activities (Spencer, 1878, pp. 630–1), or “imitations of purposive voluntary acts” (Wundt, 1884/1907, p. 357).

Much as it does today, pretending included obvious simulation of the organism's own functional behavior (“self-simulations” such as feigning sleep, fighting, or retrieval; teasing; deception), responding to nothing as if something (imaginary objects), and treating one object as if another (object substitution). Feigning (Hornaday, 1879; Romanes, 1889/1975) and “pretending” (Darwin, 1871/1896, p. 69; Wundt, 1884/1907, pp. 358) described the inhibited fighting and biting of playfighting ants, dogs, and orangutans, as well as instincts enacted out of context (James, 1890).

Understanding the origins of human symbolic behavior is an enduring quest. Little direct evidence addresses questions of how and why symbolic behavior originated, although lines of indirect evidence exist and numerous models and approaches have been offered (Lock & Peters, 1996). Among these has emerged a perspective that links our understanding of human origins with child cognitive development (Bruner, 1972; Gould, 1977; Parker & Gibson, 1979). This perspective, coupled with a focus on the evolution of cognition, has generated much recent research and scholarly attention (Bruner, Jolly & Sylva, 1976; Parker & Gibson, 1990; Mitchell, 1994a; Parker, Mitchell & Boccia, 1994; Parker & McKinney, 1999). This chapter addresses the evolution of child development with an emphasis on how its ecological contexts illuminate the evolution of a primary human symbolic domain – pretense. Further, we seek to map out connections between environmental influences on pretense and the role of pretense in the origin and persistent “reinvention” (Lock, 1980) of human symbolic culture.

Connections among play, children, the modern human mind, and culture are well appreciated. The primary adaptive aspects of our species are born within childhood, normally rich with opportunities for play and learning (Bruner, 1972), and symbolic play shares structural similarities with cultural phenomena such as myth and ritual (Bateson, 1955/1972; Turner, 1974; Geertz, 1976; Schwartzmann, 1978; Johnson, 1988). Often through pretend play (Goldman, 1998), children are transformative agents in symbolic culture (Valsiner, 1988), transforming even language (Bickerton, 1990, 1998).

Humans enjoy perceptual experiences of the world from birth, while construing of an imaginal world only becomes possible with development. Considering adult experience, Sartre (1948/1966) distinguishes two aspects of experiencing a portrait of an absent friend. If the focus is on pictorial properties – color, brush stroke, shape – the experience is primarily perceptual. If the portrait serves to evoke thoughts of the friend who is its subject, then the portrait becomes a vehicle for imaginal reflection. There is little controversy regarding the existence of these differing human experiences. However, current discussions of the topic emphasize representation as storage, rather than as a form of psychological experience. To quote a definition, “Representation is defined most simply as stored information” (Mandler, 1998). Given this climate, perhaps it is not surprising that the basic distinction between “perception” and “representation” (more accurately termed “mental representation”) is often not acknowledged in contemporary analyses (e.g., Leslie, 1987; Mandler, 1992, 1998) where infants' processing of perceived events and experience is described as “representation.” Consequently it becomes problematic to discuss distinctions between earlier infant behaviors that seem based on perceptual, sensory and motor experiences, and later behaviors that suggest a basis in mental representation, for example, recall of previous experiences which now come to mind and affect present activities. Developmental changes in representational play can mark the transition from activities based on perceptual and sensorimotor experiences to those indicating imaginal contents (McCune, 1993).

Animals are compelling presences in the world of young children, and yet the opportunities they offer for understanding topics in child development have often gone unexplored. The variations different animals present as social interactants can provide insight into a range of topics, such as children's understandings of intentionality, intersubjectivity, and sense of self-and-other. Pretense is another such topic – indeed one which draws on all of those listed above as well.

This chapter will describe the animal-role pretend play and imitation of the children in a year-long ethnographic study of a class of 24 preschoolers, ages 3 years 6 months to 6 years. At these ages, children are already adept at pretend play and we have a good understanding of their capacities for understanding others' minds. In this chapter, the focus is on how these abilities are deployed, and to what consequences, in animalrole pretend play. The chapter asks: how much and when do children pretend to be animals?; what do children do when they pretend to be animals?; and what does animal-role pretense contribute developmentally? The episodes to be described enable some generalizations about children's concepts of animal pretense; about how an animal differs from a human as a role in pretend, and about children's conceptions of animals' similarities and differences from humans. Children's animal roles entailed less use of language than did other roles, including in negotiating the pretend frame itself.