This entry reviews the main different types of play, and the kinds of developmental benefits they may bring to children. The ubiquity of play in childhood (and in most species of mammals when young) strongly suggests its benefits for development, but what these benefits are, and how important or essential they are, are still debated. Classic perspectives on the development and function of play can be found in the writings of Piaget and Vygotsky. Let us begin by tackling the issue of what constitutes play, and then turn to how it undergoes age-related changes.

Defining play

Play is often defined as activity that is both done for its own sake, and characterized by ‘means rather than ends’ (i.e., the process of the play is more important than any end point or goal). These criteria contrast play with, for example, exploration (which may lead into play as a child gets more familiar with a new toy or environment), with work (which has a definite goal), and fighting (different from play fighting as discussed later). Additional characteristics of play are flexibility (objects being put in new combinations, roles acted out in new ways), positive affect (children often smile and laugh in play, and say they enjoy it), and pretence (use of objects and actions in non-literal ways).

Main types of play

Although classifications differ, the following main types of play are well recognized: object play, pretend play and sociodramatic play, and physical activity play (exercise play; rough-and-tumble play). Of these, object play and physical activity play are seen widely in other species of mammals. Pretend and sociodramatic play are only seen in humans, apart from some possibly very elementary forms of pretence in great apes. Besides play, there is the related concept of games. Games with rules are more organized forms of play in which there is some goal (e.g., winning the game) and are not reviewed further.

Object play

This starts in infancy and may help children develop creative problem-solving skills. Researchers such as Jerome Bruner and Kathy Sylva have reported experiments with children in which they are given a chance to play with objects, then solve a task. Those with the play experience solved the task better. However, subsequent research has suggested that instruction can often be equally effective (Johnson, Christie, & Yawkey, 1999). The benefits of play need to be balanced against those of instruction, bearing in mind the ages of the children, the nature of the task, and the specificity of the learning expected – whether for specific skills or a more generally inquisitive and creative attitude.

Pretend play

This develops from about 15 months, with simple actions such as ‘pretending to sleep’ or ‘putting dolly to bed,’ developing into longer story sequences and role play (Fig. 1). Much early pretend play can be with parents, and older siblings. In Western societies especially, it is common for parents to model or ‘scaffold’ early pretend play actions. By 3 to 4 years, pretend play becomes common with same-age peers.

Pretend play among children is seen very widely in different societies. It is often imitative of adult roles (e.g., in rural societies, children may play at ‘herding cattle’ with stones and at ‘pounding maize’ with sticks and pebbles). Such play might be considered as ‘practice’ for the adult activities concerned. However, rather more ambitious developmental benefits for pretend play have been put forward.

Leslie (1987) argued that pretend play is an early indicator of theory of mind abilities. In simple object substitution pretence, the knowledge or representation that ‘this is a banana’ becomes ‘this banana is a telephone.’ Correspondingly, in theory of mind, the representation that ‘this is a banana’ is related to the representation that ‘X believes that the banana is a telephone.’ Leslie argued that this similarity suggested that pretence might be very important in theory of mind acquisition. However, this early pretend play before 3 years is often very imitative, and it is not clear whether a young child who talks into a banana is actually having the cognitive representations that Leslie describes, or is simply imitating what older children or adults do. The nature of any relationship between pretend play and theory of mind is still disputed.

Pretend play.

Figure 1. Pretend play. 

Sociodramatic play

Defined in terms of social play with others, sustained role taking, and a narrative line, this is something that children from about 3 years of age engage in a lot. Such play can be quite complex, involving an understanding of others’ intent and role, sophisticated language constructions, and the development of sometimes novel (sometimes less novel!) storylines. Smilansky (1968) suggested that sociodramatic play assists language development, cognitive development, creativity, and role taking. She also claimed that pretend and sociodramatic play were less frequent and less complex in disadvantaged children. This led her and others to develop play tutoring (intervention by an adult) to raise levels of these kinds of play; adults would provide suitable props, visits, etc. and encourage the sociodramatic play of children in nurseries and kindergartens, such that subsequently they became more able to sustain this play themselves.

Smilansky’s ideas about the value of sociodramatic play were tested by a number of experimental studies, including play-tutoring studies. In these, a group or class of children that received play tutoring were compared with those who did not. Generally, the play-tutored children improved more on measures of cognition, language, and creativity, apparently supporting Smilansky’s views.

A number of critiques were made of these studies. Many of them pointed to flaws due to selective interpretation of results, effects of experimental bias, and the use of inappropriate control groups. For example, in the traditional play-tutoring study, the play-tutored children received more stimulation and adult contact generally, so one cannot really conclude that it is the extra play that brought about the developmental benefits. Further studies took account of these criticisms. This step included balancing play-tutoring with skills tutoring (e.g., coloring, picture dominoes) and assessing outcomes blind to the child’s treatment condition. Doing so failed to reveal many differences (P. K. Smith, 1988), which suggests that benefits of sociodramatic play need not be essential for development. Nevertheless, play-tutoring does work out as equal to skills tutoring in many domains, and it is generally enjoyable and sociable for children in the preschool years, so there are sound reasons to encourage it in the nursery curriculum.

Table 1. Some criteria distinguishing play fighting and real fighting.


Play fighting

Real fighting

Facial expression

Smiling, laughing

Frowning, tearful


Kicks and blows are not hard or do not make contact

Kicks and blows are hard or make contact


Voluntarily take it in turns to be ‘on top’ or be ‘chased’

Aims to be ‘on top’ or to chase the other

How encounter starts and finishes

Starting by invitation and ending with continued play or activity together

Starting with challenge and ending in separation

One kind of pretend play, often not encouraged in nurseries, is war play, which is pretend play with toy guns or weapons, or military action figures. Many educators believe that this play encourages real aggression, though others emphasize its pretend nature and feel that no real harm results from it.

Physical activity play

This refers to playful activity involving large body activity, particularly exercise play that includes running, climbing, and other large body or large muscle activity, as well as rough-and-tumble play, that covers play fighting and play chasing. These forms of play have been reviewed by Pellegrini & Smith (1998).

Exercise play

This increases in frequency from toddlers to preschool children, peaks at early primary school ages, and then declines. Young children seem to need opportunities for physical exercise more than older children, and are more likely to get restless after long sedentary periods and to run around when released from them. Boys do more of this kind of play than girls. It is often hypothesized to support physical training of muscles, for strength and endurance, and skill and economy of movement. Another hypothesis is that exercise play encourages younger children to take breaks from being overloaded on cognitive tasks (the cognitive immaturity hypothesis). The argument here is that younger children have less mature cognitive capacities, so benefits of concentrating on a cognitively demanding task decrease after a shorter time than for older children. The ‘need’ to exercise thus helps children ‘space out’ these cognitive demands.

Rough-and-tumble play.

Figure 2. Rough-and-tumble play.

Rough-and-tumble play

This seems to increase from toddlers through preschool and primary school children, to peak at late primary age, and then decline in frequency. It takes up some 10 percent of playground time, though varying by the nature of the surface, physical conditions, etc. Boys do more than girls, especially play fighting. Rough-and-tumble play looks like real fighting, but can be distinguished from it by several criteria (Table 1; Fig. 2).

Most children can distinguish playful from real fighting, and from 8 years give similar cues to those described in Table 1. In one study, English and Italian children were found to be accurate in judging videotapes of play fighting and real fighting, irrespective of which nationality they were watching.

During the primary school years, only about 1% of rough-and-tumble episodes usually turn into real fighting, although many teachers and lunchtime supervisors think it is as much as about 30%. However, ‘rejected’ children (those disliked by many peers and seldom liked much) more often respond to rough-and-tumble aggressively (around 25% of episodes). So, it is possible that teachers or lunchtime supervisors are making general judgments about children, based on these ‘rejected’ children who maybe taking up a lot of their supervisory time.

Rough-and-tumble is often between friends. By early adolescence, however, there appears to be some change, with dominance/status becoming important in choosing play partners, as well as friendship, with a greater risk of play fights turning into real fights. It is hypothesized that rough-and-tumble play in younger children may (in addition to benefits of exercise play) provide practice in fighting/hunting skills, at least in earlier human societies. By adolescence, however, it may involve dominance relationships (e.g., using rough-and-tumble play to establish or maintain dominance in the peer group).

What do children learn from play?

Evolutionary arguments suggest that the propensity to play has been selected for, so we can expect there to be benefits to playing, and that these may vary by species, and by types of play. There can be a lot of incidental benefits to play such that it keeps children active and provides them with opportunities to encounter new situations. With human children, and with object, pretend, and sociodramatic play, there maybe a balance to draw between benefits of playing and of instruction. Instruction can be more focused on a precise goal, but play is often more enjoyable for young children and, even if less efficient for a precise goal, may foster a more generally inquisitive and creative approach to problem-solving.


Among the theoretical issues in play research remaining unresolved, two are currently especially noteworthy. The first relates to rough-and-tumble play. We know that this is primarily friendly and non-exploitative in pre-adolescents, but how does this change as children move into adolescence? Does the function of this form of play then change and, in particular, is it used for purposes related to dominance, especially for boys? The second issue relates to pretend play. An earlier phase of research queried the findings from play tutoring studies, but, more recently, pretend play has been proposed as an important component of developing a theory of mind. Greater conceptual clarity and empirical evidence are called for here, together with a willingness to learn from the problems encountered in the earlier studies (e.g., experimenter bias).

Amongst practical issues, the issue of war play continues to be debated in early education. There have been moves to ban war play in many nursery schools; however, there is also a recognition that such play may be generally harmless in itself and a rather natural play format, especially for boys (Holland, 2003). Regarding educational practice through the school years, there has been a general movement toward shortening or eliminating playground breaks. However, leaving aside social benefits of playtime, the benefits for physical activity and for providing breaks between instruction (cf., the cognitive immaturity hypothesis), argue for retaining playground breaks. More systematic study is still needed in these areas.

Next post:

Previous post: