"Rough and Tumble" Play: Lessons in Life

Evolutionary Psychology
human-nature.com/ep – 2006. 4: 330-346
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Original Article

“Rough and Tumble” Play: Lessons in Life

Pam Jarvis, Carnegie Faculty of Sports and Education, Carnegie Hall, Leeds Metropolitan University, Headingley
Campus, Beckett Park, Leeds LS6 3QS, UK, Tel: (+44) 0113 283 2600 ext. 6542, Email: P.Jarvis@leedsmet.ac.uk

Abstract: This paper focuses upon the developmental role of Rough and Tumble (R&T) play
with particular attention to the narratives that children use to underpin such activities, and to
gender differences within these. The empirical research focused upon the R&T play of children
in the early years department of a suburban primary school in Northern England. The children’s
playtime activities were ethnographically observed over a period of eighteen calendar months,
encompassing five school terms. The effects of evolution, biology and culture are recognized in
the approach taken to the analysis of data. Findings indicated that the narratives underpinning
R&T play were socially complex and highly gendered, and that mixed gender R&T play in
particular could be theorized to mirror and simplify aspects of complex, gendered adult
interaction.

Keywords: rough and tumble play, social development, evolution, culture, gender.

¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Introduction

“Rough and Tumble” (R&T) play has been defined as a physically vigorous set of
behaviors Rough and Tumble, such as chase, jump and play fight, accompanied by positive
affect from the players towards one another (Pellegrini, 1995). The play type was first
academically named as such by the anthropologist Karl Groos in his books “Play of Animals”
(1898) and “Play of Man” (1901). Much research carried out to study the social R&T play of
young human and non-human primates indicates that the activity creates valuable practice
scenarios for complex social interactions that creatures need to undertake in order to become
competent, socially mature adults (Pellegrini and Smith, 1998). Over the past 25 years, the
development of post-industrial Western society has resulted in an increasing reduction of the
time and space allocated for children to engage in such play (Pellegrini and Blatchford, 2002;
Stephenson, 2003), while over the same period, concerns about poor socialization of young
people and society as a whole have been increasingly raised; for example, a substantial
longitudinal study carried out by Collishaw, Maughan, Goodman and Pickles (2004) produced
clear evidence of a rise in conduct and emotional problems among British adolescents over the
period between 1986 and 1999. While it is likely that the issues underlying such a finding are
multi-factorial, the possible range of reasons explored by Collishaw et al. did not include
decreasing opportunities for children to engage in social free play over the period of the research.

---

Rough and Tumble Play: Lessons in Life
Modern Western education policies tend to be narrowly focused upon the development of
cognitive skills, with little attention paid to other facets of development, particularly the social.
Sylva, Roy and Painter (1980) divided play into challenging and ordinary pursuits, with R&T
very firmly in the ordinary/unchallenging category, purportedly having a low developmental
yield. Such views were taken forward into the culture of western education and consequently the
planning of many Western early-years curricula during the 1980s and 1990s. Bishop and Curtis
(2001, p. 34) quoted the Superintendent of schools in Atlanta: “we are intent on improving
performance. You don’t do that by having kids hanging on monkey bars.” However, some
researchers, typically those from the bio-evolutionary paradigm, have disagreed with such
negative views of R&T play, reflecting upon the developmental need for children to experience
independently initiated and directed collaboration with peers in order to develop the necessary
skills to become socially competent adults. Sluckin (1981) suggested that the playground is a key
venue for highly developmental play, during in which children initiate original activities,
collectively creating and negotiating rule systems. Pellegrini and Blatchford (2002, p. 62)
reflected that the playground should be considered “an extended classroom.” The lessons that the
children are learning in this classroom of the playground relate to the social and interaction skills
essential for primate adult life. Such practice activity can also be discerned in the juvenile play
activities of other primates, and to some extent, other mammalian species. This article outlines
evidence from bio-evolutionary theory and developmental research that indicates there is a
pressing societal need to provide opportunities for children to carry out collaborative free play
activities within safe environments, and that, as evolved primates, such opportunities are as
crucial to their healthy development and eventual adult competence as instruction in literacy and
numeracy.

R&T Play and Bio-Evolutionary Perspectives

Support for the greater prevalence of R&T play in males across primate species was
found by Braggio, Nadler, Lance and Miseyko (1978) in the data gathered for their observational
study comparing the behavior of children, juvenile chimpanzees and juvenile orangutans. They
found that in all three species, males undertook a higher frequency of R&T than female
conspecifics. The reason the researchers suggested for this difference is sexually selected and
hormonal; the activity of testosterone in male physiology. There is a surge of testosterone in
mammalian male bodies in early infanthood (the priming or organizing effect), then again at
puberty (the activating effect). If the priming effect is absent in males, there seem to be
corresponding behavior changes; in particular, reduced R&T has been observed in young rats
and monkeys. Introduction of testosterone to young females correspondingly creates more R&T
and mounting play. The overall conclusion drawn is that ‘male neonates have a source of
testosterone which exerts an organizing effect’ upon behavior (Whatson and Stirling, 1992, p.
107).
The importance of testosterone priming in human gender development was demonstrated
by Berenbaum and Snider (1995) and Hines, Golombok, Rust, Johnston and Golding (2002).
Congenital Adrenal Hyperplasia (CAH) is a condition in children that results from accidental
pre-natal exposure to male androgens. Berenbaum and Snyder (1995) found that girls with the
condition showed a significantly greater preference for boys’ toys and activities, while boys with
the condition did not differ significantly from a non-CAH control group in any way. Hines et al’s
(2002) study calculated the amount of testosterone present in human expectant mothers’ blood,
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-269-

---

Rough and Tumble Play: Lessons in Life
and subsequently evaluated the behavior of the resulting child at age three-and-a- half. These
researchers found that higher levels of maternal testosterone during pregnancy resulted in
statistically significant higher rates of physically active play undertaken by female offspring.
There was no correlation between levels of maternal testosterone and male offspring behavior. It
was therefore suggested that the mammalian fetus is very susceptible to the presence of
testosterone, and even slightly higher amounts than is normal in the female fetal environment
will be used to fuel a mild priming effect. This has a direct effect upon free play activities in
early childhood, resulting in more male-type behavior in the child.
A corresponding finding was made in animal research, in that female rats carried in the
same horn of the uterus as male rats showed more male-type behavior (increased mounting): “the
mere proximity of a testosterone-secreting male fetus in the uterus is sufficient to have an
organizing effect on the behavior of females” (Hall and Halliday, 1992, p. 116). Meaney and
Stewart (1985, p. 24) concluded, “play fighting represents one of the few behaviors that is
organized by early hormone exposure, but the expression of which is independent of later
activational effects of hormones.” These researchers also suggested that young male primates
rely heavily on R&T to underpin network building within their peer group, while young females
approach their social networking through grooming behavior and close association with adult
females, preparing the creature for the sexually selected roles that adult primates fulfill in natural
societies. In the case of human beings, these roles would comprise hunting, gathering and child-
raising, in which gendered patterns of interaction would exist where competition and co-
operation would be structured in subtly different ways (Bjorklund and Pellegrini, 2002).
So does such compelling evidence indicate that researchers should take a strictly
biological approach to the study of human R&T? Pellegrini (2005) proposed that if specific
behaviors occur in related species then this is an indication of shared selection processes.
However, in the highly plastic human species in particular, it is suggested that during the long
developmental period (Bjorklund and Pellegrini, 2002) the environment can have a significant
effect on the internal calibrations of a diverse set of skills, including those relating to perception,
stress coping, knowledge/skill acquisition, relationship maintenance, reaction control and the
symbols used to represent the physical and social environment (Bronfenbrenner and Ceci, 1994).
The gradual interaction of genes and environment was also emphasized by Ridley (2003),
describing how children enter the world with a particular genetic template that needs to interact
with the environment over the developmental period to produce the eventual precise settings
within the organism. Oyama (2000, p. 5) averred that development occurs by interactive gene-
environment construction rather than by reproduction of a fixed program, proposing instead that
individual human development proceeds by, “reliable genotype-phenotype correlations and
these, in turn require not genetic programs for development, but a reliable succession of
organism-environment complexes of developmental systems that repeatedly reconstitute
themselves.” Such a view of channeled biological organism development within a nurturing
environment allows us to appreciate the role of nature in human development without turning it
into a deterministic programming entity, reflecting the process that Ridley (2003) described as
“nature via nurture.”

The Role of Evolution in Narrative and Culture

The integral role that the evolved ability for abstract symbolic communication plays in
human special and individual development further adds to human developmental plasticity,
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-270-

---

Rough and Tumble Play: Lessons in Life
creating a unique role for culture within the species where abstract ideas can be shared
(Tomasello, 1999), and with the advent of literacy, built upon by subsequent generations.
Evolution has thus provided the human primate with language; but how might the abstract
symbols used in such communication be channeled through our evolved biology? The concept of
narrative deals with how human beings construct cohesive linguistic accounts rooted in shared
cultural understandings; might human narrative constructs be mediated by evolutionary factors?
Bruner (1990) referred to this as the “biology of meaning,” proposing that while such a concept
might initially appear to be an oxymoron, it is a legitimate and highly salient research category.
Human beings are creatures who are evolved to critically rely upon sharing symbolic meanings
to live in their world, and such symbolic meanings, “depend upon the human capacity to
internalize language and use its system of signs…such a social meaning readiness is a product of
our evolutionary past” (Bruner, 1990, p. 69). Human beings may understand many, sometimes
overtly similar aspects of their world very differently, depending on the story or “fabula” that
they attach to them, highlighting the ways that human beings create, “products of the mind,
build[ing] them into a corpus of a culture” (Bruner, 1986, p. 45). Lyle (2000, p. 55) proposed
that human beings inhabit a, “largely story shaped world…[thus operating as a] storying animal,”
making sense of physical and social environments via stories and narratives. An emergent
question is therefore: might the genders construct rather different stories, originating from a
sexually selected, evolved neuronal template, and consequently, might the earliest expression of
such narratives be represented in the first free play activities?
In the pursuit of such extension of knowledge, perhaps Gilligan’s (1993) concept of
gendered “voice” can be proposed to be a viable window to the human, gendered template in
interaction with the environment, an external manifestation of how the, “genes build the
organism and its instincts through a flexible process of development” (Ridley, 2003, p. 129).
Studies of playground activity typically find that boys are more involved with physically active
play that revolves around issues of dominance and status, while girls prefer more sedentary play
exploring more symmetrical, co-operative relationships (Maccoby, 1998; Pellegrini, 2005); in
order to initially investigate the correspondence between linguistic and behavioral gender
differentiation, there needs to be some exploration of whether there are correspondences in the
different types of play undertaken by girls and boys and the ‘voices’ that they use to narrate and
describe their play activities. Kyratzis (2000) summarized the current early years gendered
language research evidence as follows: girls build a sense of community in their language. Their
conversation indicates that they are concerned with being nice, and creating intimacy and
solidarity within their friendship groups, wishing to be seen by their friends as moral and
loveable. People who are perceived to be mean are excluded from the group. Boys are
concerned with being adventurous, risk taking and flouting authority outside the friendship
group. They do not seek to appear nice, but they do have underlying concerns about the
cohesion/solidarity of the group. People who are weak are seen as not worthy of male group
membership.
It may initially seem from such a description that girls within female friendship groups do
not compete; this was the position taken by Sheldon (1990), who proposed that boys’ groups are
adversarial and girls’ groups affiliative. However, Kyratzis (2001, p.4) firmly disagreed with this
position, proposing: “our views on conflict are andocentric and fail to acknowledge girls’
assertion.” In an earlier paper, she emphasized these subtle differences between the gender
“voices,” proposing that both genders vie for position in the peer group, boys seeking to be the
most dominant, and girls the nicest. Girls tell stories to indicate and consolidate alliances, while
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-271-

---

Rough and Tumble Play: Lessons in Life
boys’ stories are designed to emphasize to one another how naughty (authority flouting/
dominant) they are: “narrative is a political activity, serving to establish political alignments by
talk” (Kyratzis, 2000, p. 278). Crick (1996) argued that the reason that some researchers propose
that girls are not aggressive is because they are focusing on the wrong type of aggression in their
research methodologies. While girls do not typically engage in direct physical or verbal
aggression, they do employ relational aggression, undermining other children’s relationships
within the peer group. Roy and Benenson (2002) linked this gender difference to a sexual
selection explanation: adult males can achieve their maximum chance to produce offspring by
directly competing with other males for status and resources, which enhances their attractiveness
to females as superior providers. However, the adult female can produce less offspring and must
invest far more of her physical resources in each child than the male parent. In the natural
environment, primate females typically care for their children within female kin groups; as such,
the pathway to successful reproduction for the female is through building and maintaining strong
relationships with the other females in her kin group. Consequently, covertly undermining other
females within the group whilst simultaneously maintaining a pleasant façade would appear to be
a highly adaptive strategy for females, allowing them to undermine specific competitors whilst
maintaining good relationships with the majority of the group. Charlesworth and Dzur (1987)
found evidence of such co-operation mixed with manipulation amongst 4-5 year old female, but
not male groups. Their study of single-gender interaction found that girl groups tended to form
under the control of a single, dominant female who used relational aggression to retain her
authority, while male interaction tended to involve dominance behaviors from the majority of
group members. While some boys would be more successful than others in their dominance
attempts, these researchers noted that all-male groups did not typically fall under the control of a
single boy to such a great extent.
The empirical findings of Kyratzis (2000) would also seem to support this model of
gendered interaction. She described men and boys using a competitive style in their
conversations, depicting a world in which individuals are engaged in contest, while women and
girls are more likely to use the conversational style of double voice discourse, a highly assertive
conflict speaking style, which nevertheless uses mitigating language content in an apparent
attempt to quell discordance amongst the social group (Kyratzis, 2001). Girls’ talk thus has a co-
operative surface structure but provides a framework where they can compete for emotional
dominance within their social group, while boys talk has a competitive surface structure but
provides a framework for companionship and group solidarity. It can therefore be proposed that
co-operation can become a very effective form of competition among female groups
(Charlesworth and Dzur, 1987), and that within each gender cohort, “narrative manages power,
conflict and social ranking within friendship groups” (Kyratzis, 2000, p. 295). Hence there is
evidence to suggest that the way that narratives are typically used by each gender can be
theorized to be to some extent dependent upon the evolved, gendered bio-psychological template
of the individuals concerned.
An empirical example of such gendered interaction can be found in a study carried out by
Marsh (2000). She invited children within a nursery setting to undertake active, fantasy play
within a “Batcave” role play area, emphasizing that both boys and girls could play and be
“Batmen” or “Batwomen.” It was clear throughout the research period that there were distinct
differences between the gendered superhero discourses. “Batwomen were most likely to rescue
victims, often those who were clearly vulnerable such as children, maintaining good
relationships with their fellow Batwomen, the Batmen chased and captured villains” (Marsh,
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-272-

---

Rough and Tumble Play: Lessons in Life
2000, p. 218). The evolved, gendered template therefore also appears to be reflected in the voices
described in this study, girls creating imaginary narratives where good relationships are
maintained, compared to the male-generated scenarios in which boys explored status and
dominance.
A “warrior discourse” among boys was additionally identified by Jordan (1995, p. 76).
She reflected, “we have, as far as I know, little in the way of explanation of how or why these
narratives gain such a grip on little boys, but the evidence that they do and have done for
generations is inescapable.” It could be suggested that the theoretical missing link in this debate
may be the evolved gender template postulated above. In conclusion, there is a body of empirical
evidence to suggest that the genders are influenced by their underlying biopsychology to build
gendered play narratives.
Consequently, it can be proposed that there is thus a distinct emergent dichotomy with
regard to the study of human R&T, indicating a clear link and a distinct difference between the
R&T play of children and juvenile non-human animals:


The basic occurrence of R&T play in human children can be shown to have clear
evolutionary roots in the non-verbal play of earlier species. This has been demonstrated to be
mediated by evolved, sexually selected differences, underpinned by the complex role of
testosterone in male mammal physiology.

Such play in human beings is likely to show greater variability and complexity than that
observed in animals due to human developmental plasticity and the human ability to incorporate
complex symbolic products of culture, narrative and imagination into play actions. Such
narratives can be theorized to differ between males and females, as a reflection of different,
sexually selected bio-psychologies.

“Deep” Free Play: The Basis of Primate Social Learning

When examples of primate collaborative free play are considered, it can be seen that they
form ideal practice scenarios for the young creatures concerned to begin to explore issues that
arise within the complex social lives of primate adults (Meaney and Stewart, 1985). This type of
play naturally occurs in many mammalian animal species, most particularly those with adult
societies that center around complex social networks. For example, Fagan (1976) described play
fighting among rhesus macaques where the animals play-bite and feign injury; Bertrand (1976, p.
320) described monkeys playing “tag” where, “one monkey approaches another with play leaps
or … play face, hits it and runs away in order to be chased” and Biben (1998) described the play
of juvenile squirrel monkeys during which running, jumping and swinging from branches
involves much play fighting, the results of which impact upon the dominance hierarchies within
the group.
Human beings have the most complex social networks found among naturally evolved
creatures, and our social free play has an important extra dimension, that of language. O’Donnell
and Sharpe (2004, p. 90) proposed that boys’ perceived peer groups create R&T play situations
in which they explore “a sense of nationalism and territory.” This finding neatly illustrates both
the singularly human reliance upon linguistic communication, in which the construction and
communication of highly abstract ideas emerge, such as the concept of “nationalism,” and the
human similarity with other mammalian species who also use physically expressed play to
explore dominance hierarchies and learn to deal with complex sets of vertical and horizontal
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-273-

---

Rough and Tumble Play: Lessons in Life
social relationships. Human beings alone use rich imaginative models rooted in language to form
abstract concepts of their environment and their place within it (Carroll, 2004), which, during the
developmental period, are combined with physical play interactions that show great similarity to
the highly physical play of non-human animals.
In the past, Western children had many everyday opportunities for such “deep” or
“serious” play (Bruner, 1976, p. 57), exploring complex social primate relationships in out-of-
school independently directed collaborative activity, involving larger kin groups interacting in
public areas principally used by pedestrians rather than motor vehicles. However, during the last
quarter of the twentieth century, western children have increasingly inhabited vehicle-clogged
areas, within singleton or two-child nuclear families, where they are driven to school and not
allowed to play unsupervised in outdoor environments due to parent perception of traffic and
“stranger” danger (Stephenson, 2003). Much early twenty first century out-of-school activity
involves armchair-bound, solitary playing of computer games, some of which re-create physical
activities such as ball games and skate-boarding, complete with digital companions whose
reactions have been artificially programmed by adults. More physically active social pursuits are
increasingly likely to involve being driven to various supervised venues for adult-directed
activities. Corsaro (1997, p. 38) described this process as, “the institutionalization of more and
more children’s leisure activities.”
The only regular chance that many contemporary Western children therefore receive to
engage in interactive free play is within the environment of the school playground; however, in
England, free play time for state school pupils has been continually reduced over the past fifteen
years by rearranging the school day around a shorter lunch period and removing the afternoon
break (Pellegrini and Blatchford, 2002). When the current debates over the highly didactic
English National Curriculum (and other similar western school curriculums) are also considered,
it can be proposed that many western children are placed within highly adult-directed
environments with little time or space for ‘deep’ interactive free play.

R&T Social Free Play: The Research

The majority of previous research focusing upon R&T play has been carried out to
investigate the physical aspects of boys’ R&T play (e.g., Coie, Dodge, and Coppotelli, 1982;
Dodge, Coie, Pettit, and Price 1990; Pellegrini, 1993a; Pellegrini, 1993b; Pettit, Bakshi, Dodek
and Coie, 1990). However, the research undertaken by the author focused equally upon both the
play of both genders, with the purpose of investigating gendered narratives in R&T and chasing
behavior. The participant establishment was a medium size primary school located within the
suburbs of a large city in Northern England. It was selected for its relaxed attitude to playground-
based R&T play, as it was the researchers’ opinion that a school with a more structured approach
to playtime (recess) might offer little opportunity to observe children in authentic free play. The
participant school had an integral nursery class, which children attended from the September or
January following their third birthday, moving up into the Reception class of the main school in
the September following their fourth birthday, which is the conventional English practice.
I first met the children who were going to become my research participants in April 2002,
towards the end of their nursery year. The principal participants of the research were nine girls
and nine boys, born within the six months between September 1997 and April 1998. There was
some additional emergent participation from children with whom this focal sample engaged in
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-274-

---

Rough and Tumble Play: Lessons in Life
play within their school playground, and from adults engaged in the daily activities of the child
sample.
The research was undertaken in an ethnographic, broadly participant observational
fashion. I visited the children in nursery, arranged the necessary ethical permissions and carried
out preliminary observations during their final nursery term between April and July 2002, during
which time they got used to my presence and the experience of being observed by an adult
speaking quietly into a small dictaphone. The set of observations used as data for this piece of
research were subsequently carried out between September 2002 and November 2003. The
children were placed in the Reception class of the main school between September 2002 and July
2003, moving up to Year One (first grade) in September 2003.
I used the approach of modeling my interaction role with the children as much as possible
upon a volunteer parent-helper in school, attempting the “observer as participant” research
methodology (Banister, Burman, Parker, Taylor, and Tindall, 1994, p. 39). I responded to
children’s requests for help with buttons and shoelaces, and overtures to ‘show’ objects, while
avoiding involvement in any of the directive or play-theme generating adult roles within the
playground and the classroom. My gender, age and previous life experience were probably
helpful in this respect; in age and appearance I was a fairly typical example of a mother of
children within this age group, and added to my professional role as a part-time teacher in a
different establishment, I also had previous experience as a volunteer parent-helper at my own
children’s primary school. After a very short period of initial interest from the children (2-3
weeks) I was treated by the sample as part of their usual classroom adult cohort, being asked for
help with various everyday practical tasks, occasionally shown objects of interest and otherwise
fairly generally ignored.
I never approached the children on disciplinary matters, and where children initially
asked me to referee arguments or deal with disciplinary issues I referred the complainant(s) to
other relevant adults. After a few weeks I found that the sample and their classmates did not tend
to bring these matters to my attention, or (as far as I was aware) avoid or hide minor behavior
violations when I was present. I felt that I had taken a more naturalistic approach than that used
by Blurton Jones (1967) who carried out observations of 3-4 year olds in a nursery setting during
1963-4, reporting that his technique was to be as unresponsive to his child participants as
possible. I did not find, as reported by Smith and Connolly (1972) that the frequency of the
children’s approaches increased when I carried out basic routine interaction with them; however
that might be due to my age and gender, as in the studies that Smith and Connolly referred to the
researchers were all male, while, in similarity to the environment I was researching, the school
staff were all female. Smith and Connolly reflected on the possibility that female observers
might elicit less curiosity from the children, and my experience seemed to bear this out, possibly
aided by the fact that several of the adults with whom the children regularly interacted in their
nursery and reception class environments were volunteer parent (or more accurately, mother)
helpers who frequently worked with small groups or individual children on craft or reading
activities.
The principal technique used for the observations was that of “target child” (Sylva Roy
and Painter, 1994, p. 9). The final total of target child observations undertaken was seventeen
male and sixteen female target child observations, two observations of fifteen of the focal group,
and one observation each of the remaining three. Each child who was the subject of two
observations was observed once during a playtime (recess) period (20 minutes) and once during
a lunchtime period (approximately 40 minutes, depending on how quickly the child finished his/
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-275-

---

Rough and Tumble Play: Lessons in Life
her lunch and emerged into the playground). I also carried out 5 “target area” observations
(Sylva et al., 1994, p. 9), focusing on the regular soccer play that took place on the grassed areas
during the summer term, mainly involving the Reception and Year One (equivalent to first
grade) boys. My final six observations were carried out as “theoretical sampling” (Strauss, 1987,
p. 39) over the autumn term September 03 – November 03, directing the investigation flexibly
towards specific R&T play that occurred among the sample, rather than focusing whole
observation periods towards the interactions of individual children or interactions within specific
areas of the playground.
My usual procedure was to walk around the playground dictating my notes quietly into a
dictaphone, standing approximately 10 yards away from the relevant area or the relevant child
and his/her playmates. If a child reacted by stopping his/her activities and looking directly at me,
I would walk away for a few moments and look elsewhere, returning when the child(ren) were
reabsorbed in play. This very seldom happened after the pilot period. When children became
very absorbed in play I was usually able to move close enough to hear some of what they were
saying; I also made a practice of chatting to them about what they had been playing during that
play period as they walked towards their class lines after the bell had gone. I dictated my field
notes in an ethnographic style, describing all the target child’s play activities, and associated
language during the period of the observation, noting any interruptions, then fully transcribed the
tapes. Where I carried out target area observations, I focused on the interactions of the children
within the focal sample who were engaged in play within the area. A typical interruption during
my observations was caused by the fact that children were allowed to go indoors to use the
toilets during lunchtime breaks. I found that this sometimes caused an interruption of individual
children’s outdoor play for up to 10 minutes, particularly with respect to the girls who tended to
chatter in the hand washing area directly outside the toilet cubicles. Pellegrini (2005) also noted
that girls were more likely than boys to find strategies to avoid being outdoors during stipulated
outdoor play periods.
Each observation transcript was summarized onto an observational data sheet. I then
organized the summarized data into gender-based groupings with sub-sets for girls’ play, boys’
play and mixed gender play. Using these references, I returned to the full set of field notes and
collated all the information relating to the children’s “scripting” of their R&T play under the
headings of girls, boys and mixed gender play. I found that several such scripts could be divided
into over-arching themes (for example, boys chasing/girls fleeing) with more detailed stories
being attached to specific episodes of play, which varied from play session to play session.
These had a certain pragmatic quality in that the specific story tended to be tailored to the
available play environment at the time. For example, in hot, dry weather, part of the process was
likely to involve children lying on the grass for short periods of time to signify that they were
“out” of a game, while in wet weather the process was changed so being “out” was signified in a
different way, for example, standing against a wall. This was predicted by the findings of
Pellegrini, Huberty, and Jones (1995) who found that temperature creates a clear effect upon
children’s play activities. In order to define an over-arching theme underlying a particular story I
deemed this a “narrative,” while the specific story was called a “fabula,” the latter term being
taken from Bruner (1986, p. 45). Most of the R&T play observed appeared to have some aspect
of narrative that engaged the children and directed their play, in the sense that the moment-to-
moment activity involved had a specific meaning for the child or children concerned.


Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-276-

---

Rough and Tumble Play: Lessons in Life
Analysis: Single Gender Play

The rarity of girls only R&T play is the first aspect to remark upon; only four
observations from the set of thirty-three target child observations contained girls only R&T play.
One of these generated (arguably) the most complex and original fabula observed scripting an
episode of R&T play, the story of a witch and a magic rabbit. Another girls-only R&T game was
quite lively and involved a large amount of physical contact. It involved spinning quickly around
a signpost and then lying on the grass, which at one point, involved a gentle pile-on as the girls
laid on top of one another, giggling and hugging. The underlying fabula for this play episode
appeared to be “putting baby to bed,” in that baby gets up, goes and plays (spins around the post)
and then gets tired and fractious so has to go to bed again.
Fifteen episodes of boys only play were observed during the target child observations; ten
of these involved highly active R&T. In single gender R&T the boys tended to rely on current
media for fabulas. Beyblades, a Japanese fantasy cartoon about spinning warriors was very
popular at the time of the observations, and several boys’ spinning activities were observed
where they pretended to be the Beyblades, sometimes taking on the names of the characters. The
game involved spinning while karate chopping at ones’ opponents, the aim being to knock the
other player out of a “ring.” When a player was knocked “out,” he generally collected himself
and went straight back into play; there did not seem to be any concept of being “out” for any
length of time. It was this higher level of energy in boys’ only play that separated the genders
most distinctly. A sustained all-boys’ “Robot Wars” game was also observed, based on a factual
television program featuring fighting robots. On the face of it, this was chasing and catching
play, as was the “witch and magic rabbit” all-girls game. However the pace, roughness and
particularly the nature of the contact between the players indicated subtly different gender
orientations to chasing activity. Play punching/karate chopping was very evident in the Robot
Wars game; a specific chopping motion was later described to me by one of the players as the
action of “Mr. Psycho’s hammer.” This type of movement was not observed in the witch and
magic rabbit game, which principally involved the child playing the rescuer hugging the child
playing the rabbit to protect it from the child playing the witch’s malevolent touch, which, the
players agreed, would turn the rabbit into wood. There was also a clear tendency to hug and
cuddle in the brief and gentle girls’ pile on, the underlying fabula for this being rather caring/
maternal: “putting baby to sleep.” The more common boys’ pile-ons observed, for example
connected to a ‘cheetahs and leopards’ fabula, were conversely related to brawling and play
strength competition. There was more direct physical confrontation in most of the boys’ games,
and less coherent vocalization than was observed in girls’ play.
At times a clear hierarchy seemed to be in operation within the boys’ play that did not
appear so obviously within girls’ or mixed gender groups, which was indicated by evidence of
boys ‘passing on’ feelings of subordination to other, often physically smaller boys, usually in
more overbearing play fighting behavior with particular individuals. Competing male claims of
“toughness” were also routinely made; for example, one boy declared during soccer play: “I’m
one boy but I can tackle a thousand men.” Another, physically smaller boy replied rather
uncertainly, “I can tackle lots of men.” The name of the very popular and glamorous (clearly
?ber-male) England soccer captain was additionally frequently raised by the boys to proudly
boast of a successful outcome in the soccer games; for example a typical comment from a boy
scoring a goal to the other players was “look – like Beckham!” The vocalization was also
sometimes accompanied by a well-observed imitation of Beckham’s characteristic “victory
Evolutionary Psychology – ISSN 1474-7049 – Volume 4. 2006.
-277-

---