The Arts in Psychotherapy 37 (2010) 5664

Contents lists available at ScienceDirect

The Arts in Psychotherapy

Stimulating creative play in children with autism through sandplay

Lucy Lu, M.A. a , Fiona Petersen, B.SW, DESS. a, , Louise Lacroix, M.A., ATR a,b , Ccile Rousseau, M.D. a,c
a

Youth Mental Health, CSSS de la Montagne (CLSC Parc Extension), 7085 Hutchison Room 204.2, Montreal, QC, Canada H3N 1Y9
Department of Creative Arts Therapies, Concordia University, 1455 de Maisonneuve Blvd. West, SVA 264, Montreal, QC, Canada H3G 1M8
c
Division of Social and Cultural Psychiatry, McGill University, Canada
b

a r t i c l e

i n f o

Keywords:
Sandplay
Play
Autism spectrum disorders
School-based interventions
Action research

a b s t r a c t
A school-based action-research intervention with children with autism spectrum disorders investigated
whether sandplay could be used as a medium to stimulate creative and symbolic play. Twenty-ve elementary school children in four separate special education classes within the regular school system
participated in sandplay workshops once a week for 10 sessions. The intervention aimed to stimulate
communication, social interaction, and symbolic play through the use of rhythm- and movement-based
rituals and sandplay. Over the 10-week program, children demonstrated through sandplay increased
verbal expression, engaged and sustained social interaction, and increased symbolic, spontaneous, and
novel play. The study suggests that creativity-based interventions provide a complementary approach to
behavior/social skills-based intervention models prevalent in schools working with children with autism
spectrum disorders.
2009 Elsevier Inc. All rights reserved.

Introduction
The current practice of education of children with autism and
pervasive development disorders (PDD), or more generally autism
spectrum disorders (ASD), in Canada and more widely in North
America, has largely focused on integrating children into the mainstream education system, whether in special education classes in a
regular school setting or individual children integrated into regular
class settings (Brock, Jimerson, & Hansen, 2006; Bryson, Rogers, &
Fombonne, 2003; Hess, Hein, Morrier, & Michelle, 2008). Schoolbased education approaches for children with ASD demonstrate
a propensity towards highly structured behavior-based interventions, which have proven effective in addressing the core decits
in communication, social interaction, and restricted repertoire of
behaviors, activities and interests (Brock et al., 2006; National
Research Council, 2001). Although structured learning is essential
for this population to learn functional skills, proponents of developmental approaches contend that behavioral approaches are limited
in encouraging natural interpersonal interactions and spontaneous
symbolic play (Greenspan & Wieder, 2006, 2007). As spontaneous,
exible, imaginative, and social qualities of play are typically underdeveloped with children with ASD (Wing & Gould, 1979; Wolfberg,
1996), interventions that can promote this quality of play could be
benecial to children with ASD.

Corresponding author.
E-mail addresses: lucyluAT@gmail.com (L. Lu), ona.p@sympatico.ca
(F. Petersen), louise-lacroix@videotron.ca (L. Lacroix), cecile.rousseau@mcgill.ca
(C. Rousseau).
0197-4556/$ see front matter 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.aip.2009.09.003

The gains for children with ASD in developing their ability to

play is multifold as it is suggested that play provides a medium
through which children develop cognitive and emotional skills,
exible thought, and experimentation with roles, as well as the
opportunity to interact with others and to express themselves
(Wolfberg, 1996). Structuring and enhancing play with children
with ASD gives them a sense of mastery, as well as increasing
their pleasure and their motivation to play, which is a justiable
aim in itself (Boucher, 1999; Wolfberg, 1996). There is a growing
body of applied research (Jarrold, Boucher, & Smith, 1996; Jarrold,
2003; Libby, Powel, Messer, & Jordan, 1998; Sherrat & Donald,
2004; Sherratt, 1999, 2002; Wieder & Greenspan, 2003; Wolfberg
& Schuler, 2006; Wolfberg, 1996) dedicated to understanding the
obstacles that children with ASD face in initiating and producing
spontaneous and imaginative play, devising means to support and
enhance it, and ultimately to help children with autism partake in
their peers culture where play is of such central importance.
Sandplay, the central tool of this program, as a creative medium
is particularly adaptive to many different populations and settings.
Inspired by Margaret Lowenfelds World Technique, Dora Kalff, a
Jungian therapist, developed Sandplay in the 1940s as a non-verbal
therapeutic modality to work with children (Kalff, 1980/2003).
Within the free and protected space provided by the therapist,
a child or adult creates a concrete manifestation of his or her inner
imaginal world using sand, water, and miniatures objects. Thus,
sandplay illuminates the clients internal symbolic world and provides a place for its expression within a safe container, the sand
tray (Friedman & Mitchell, 2008, p. 1).
Many adaptations have since been created (Mitchell & Friedman,
1994) including sandtray therapy that has been adapted to group

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

situations and allows each participant the opportunity to create

their own world while in the presence of others, with the possibility of sharing their experience with the group (De Domenico,
1999). De Domenico (1999) reports that sandtray therapy in a
group context helped improve self-esteem, decrease isolation and
withdrawal or extreme introversion, and increase curiosity and
self-reection. Sandplay has been used with children in individual therapy (Mitchell & Friedman, 1994; Van Dyk & Wiedis,
2001), and as a way to work therapeutically and preventively
in school populations (De Domenico, 1999; Kestly, 2001; (as
sandplay without therapy) Lan, 2008; Pabon, 2001), and with
school-aged immigrant and refugee children to promote adaptation through creative expression (Lacroix et al., 2007; Rousseau,
Lacroix, Singh, Gauthier, & Benoit, 2005). This type of creative
group sandplay activity within a class in a school setting is a
means to promote imaginative play and symbolic expression as
well as facilitating communication. The focus is more educational and developmental than therapeutic; however, therapeutic
effects may be apparent during the creative activity. Literature
on sandplay populations with ASD is sparse (Tanguay, dAminico,
Dolce, & Snow, 2004) and therefore needs further investigation
to explore its potential as a tool to work with children with
ASD.
Children with ASD are often grouped in the same classroom
despite displaying a wide range of social and cognitive abilities.
Sandplay employed as a creative group sandplay activity presents
the potential to stimulate play at multiple developmental levels from tactile exploration to early and more complex social
play, from functional play to symbolic representation, and pretend play. Also sandplay does not completely depend on verbal
expression, but does promote communication and shared meaning through the storylines developed in the play. In all these
respects, sandplay has the potential to be adapted and incorporated into existing educational practices with children with
ASD.
As ASD populations demonstrate diverse learning needs, school
policies in education advocate for personalized programming
(Bryson et al., 2003), which will often incorporate a variety
of treatments such as behavioral, developmental, and occasionally creative-based interventions within a structured educational
framework (Hess et al., 2008). It has been noted that interventions are often modied and altered in the actual school setting
(Stahmer, Collings, & Palinkas, 2005), suggesting that although an
intervention has been proven effective in a controlled research
setting, teachers need to nd ways to incorporate the myriad intervention practices to the actual education settings when working
with children with ASD. Action research provides an avenue to collaborate with educators, adapting the intervention as it is being
studied (Stringer, 2007), in order to see how creative interventions
can be integrated with and complement existing practices in the
actual setting that address the core decits of children with ASD.
This preliminary study describes an action-research approach to
investigate and ask the following questions:
1. How does sandplay adapted as a semi-structured creative intervention support the developmental skills of children with ASD
in the areas of communication, socialization, and symbolic elaboration?
2. How do children with ASD with different levels of impairment
respond to the sandplay intervention?
Description and rationale of intervention
Each session took place during a 60 min class period and
consisted of an opening ritual, a sandplay period, a storytelling

57

exchange, and a closing ritual. The intervention was designed in

such a way that it provided a structured routine, while also providing the framework for spontaneous and child-directed play and
expression.
Opening ritual
Opening and closing rituals provide children with the consistency of a predictable beginning and clear ending that is reassuring
and delimits the intervention from the regular school routine
(Martin, 2001). The opening ritual was a brief 510 min activity
that welcomed the children into the creative space using physical,
verbal, and imaginary play as warm-ups to the sandplay activity.
Ribbon sticks and egg shakers, encouraging gross and ne motor
movement and rhythm, were used in brief activities involving
mirroring, the naming of personal tastes and feelings, play acting, and collective storytelling. The opening rituals were designed
to enhance affect and encourage sharing so that children could
express and interject their ideas and create gestures that we could
all imitate. Some examples included imagining being different
animals, eating different favorite foods, reliving the Montreal Canadiens hockey win, or taking a camping trip together. Each ritual
was modied to the developmental level and verbal capacity of
each group of children.
Sandplay
The main activity of sandplay was adapted as creative group
activity to address the play needs of children with ASD. Whereas
Sandplay encourages a passive, containing role for the therapist
(Lowenfeld, 1979), we practiced a semi-structured child-centered
approach, employing techniques like narrating the childs play,
mirroring, and offering gurines to draw out the play, while
remaining sensitive to the childs reactions (see Knoblauch, 2001).
This type of relational approach to play is based on principles of
the Floortime model (Greenspan & Wieder, 2006), that encourage observing, following the childs lead, opening and closing
circles of communication, so that children can elaborate on their
self-initiated gestures in order to expand on their ideas and
imagination. The aim of using a relational approach in this intervention was to help support the development of imaginative
play.
Although the sandplay intervention was adapted for the school
context, the main identiers of sandplay were utilized. A sandtray,
according to Kalffs dimensions (1980/2003), was provided for each
participant as well as bins grouping different gurines and objects
that the children would share. In sandplay, one can choose from
a vast array of miniature human and animal gures (both realistic
and fantastic), trees, transportation, housing, and food, as well as
miscellaneous objects such as marbles and feathers that provide
color, textures and material for original constructions, to create
scenes in a sand tray (Kalff, 1980/2003). The inside of the sand
tray is painted blue, so it can easily represent the sky or a body
of water. Kalff (1980/2003) insists on trays of a very specic size,
corresponding to the childs central eld of vision and facilitating
safe immersion in play. The frame is also purported to have a limitsetting function (Carey, 1990) and acts as a regulating, protective
factor (Kalff, 1980/2003). The sand provides a soothing medium
that stimulates the sense of touch, smell, and sight and is often
associated with playing in the park or on beaches (Lacroix, 2002;
Steinhardt, 1998). Small quantities of water were offered as an additional tactile dimension to the sand quality and played a role in
attracting and sustaining childrens attention. The different tactile
qualities of the sand and the diversity of gurines and objects stimulated imaginative play, which eventually moved to the process to
storytelling.

58

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

Storytelling
A storytelling period followed the sandplay activity for the more
verbally expressive groups of children. For the younger and less
verbal groups, limited verbal exchanges would be one-on-one with
the teachers or therapists, or occasionally children were invited
by the therapists to walk around and view the sandplay images
of their classmates, and to make one- or two-word descriptions
of their favorite objects. The more verbally expressive groups of
children were able to sustain enough attention to view and listen
to each of their classmates, as they either enacted or told their more
developed stories. In some instances, a few of the children chose to
work together to make collective sandtrays and stories.
The participatory nature of action-research encouraged the collaboration of teachers to modify the intervention to meet the
specic developmental and academic needs of the children. For
example one teacher structured the storytelling by asking the children to name the time, place, plot, and ending to their stories, while
another teacher contained the activity by holding up a 2-min timer.
The children appreciated the structure as it formalized the activity, providing them the undivided attention of their viewers. Many
children who appeared unfocused during their play made up stories
on the spot for their teachers and peers to appreciate. Regardless
of the changes or modications made to this storytelling section,
each class ended with the same closing ritual.
The closing ritual
The closing ritual remained constant and consisted of hand
clapping that mimicked the beginning, climax, and ending of a
rainstorm. Excess energy from the activity allowed an emotional
release through the dynamic physical movements at the climax of
the storm, while the slow mimicking of the patter of rain provided
a calming group experience that permitted transition back to the
regular school program. The children often requested to direct the
familiar activity themselves, which led to unforeseen offshoots of
the ritual, such as changing the rhythm, or adding different gestures or sounds, or adding their own personal touches such as the
sun coming out in the end. The closing ritual acted as a predictable
container to the creative space and, when it was child-directed,
it promoted self-esteem, group awareness, and the creative and
emotional investment that enhanced its appeal for the children.
Method
Action research
Following on previous research using sandplay adapted to a
school setting (Lacroix et al., 2007), in this research two art therapists implemented a creative intervention using sandplay and
interactive symbolic games. Action research is based on the premise
that research be conducted on real problems or issues faced by
practitioners in order to develop localized solutions (Stringer,
2007). In this case school teachers and psychologists working with
children with ASD in an integrated school setting expressed the

need to implement a more diverse range of therapeutic and educational practices that would meet the needs of children with
ASD. This research objective aimed to investigate whether creative interventions, such as the use of sandplay, could be used as a
semi-structured pedagogical tool that teachers could appropriate
in order to promote communication, socialization, and the capacity
to symbolize.
Population and setting
The children were a part of four special needs classrooms in
an elementary school setting in Montreal, Canada, designated for
children who met standardized criteria for autism and PDD. Each
classroom had 67 children, with one teacher, one special needs
educator, and occasionally a special education intern. The normal
ratio of adults to children in the class was approximately 13, while
during the intervention the ratio was 23.
Children participating in the program ranged from the ages of
712 years old, with an average age of M = 9.9 (N = 25), of which
23 were boys and 2 were girls (Table 1). Children were assigned
to their class according to age and developmental level. However,
even within these assignments, there was a large range of developmental expression within each class. In addition to the diagnosis
of autism or PDD, some children also demonstrated delays in
motor coordination and impaired hearing affecting elocution. This
French-speaking school was representative of an ethnically and linguistically diverse school setting in Montreal and, despite the small
numbers in these classes; the composition of the special needs
classes reected this diversity.
All parents consented to permit their children to participate in
the intervention and research. Most of the children participated
in the entire length of the 10-week program, with some absences
among six students for one to four sessions due to occasional sickness. On average, there was a 95% attendance of students to all of
the 10 sessions.
Data collection
Qualitative data were collected after each session following an
observation grid based on symbolic developmental levels of play by
Greenspan (Greenspan & Wieder, 2006) and grids of former sandplay projects by the team (Lacroix et al., 2007) (see Appendix A).
Observations of each child were noted for receptive and expressive
communication, social interaction, and symbolic expression during
all phases of the intervention. Children left the images intact after
each session and the therapists cleared the tray before the next
group. Photographs of each nal image of the sand tray were taken
along with particular reconstructions of images that were pertinent to the storytelling. As each child differed in their capacity to
express and communicate verbally the stories they created, observations were noted of their affect during play, level of engagement
with the materials, social interactions, and any verbalizations or
stories they made about their sandplay.
Teachers completed a questionnaire about their expectations of
the program and observations of the childrens symbolic capacity

Table 1
Categories of creative expression.
Creative expression

Sub-categories

(1) Pre-symbolic expression

(1.1) Tactile or sensorial exploration

(1.2) Functional play

3
4

3:0
4:0

89
812

8.7
9.8

(2) Symbolic expression

(2.1) Ritualistic play

(2.2) Beginnings of story-telling
(2.3) Symbolic themes organized into story form

6
7
5

5:1
7:0
4:1

710
712
1012

9.3
10
11.2

712

9.9

Total number of children (N)

Number of children (n)

25

Gender ratio of boys:girls

Boys n = 23; girls n = 2

Age range

Mean age (M)

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

before the 10-week intervention. The two art therapists and the
psychologist met with the teachers at the beginning, middle, and
end of the program to gain feedback, share observations, and to
discuss the program. At the end of the program, a focus group was
held with the teachers by another member of the team.
Analysis of ndings
A content analysis (Stringer, 2007) was made of the data collected from the observation grids, the images of the sandtrays, and
the teachers observations. An initial global analysis of the program
was noted for the level of engagement of the children in the activity, the capacity to engage symbolically and the complexication
of stories over time. In the second level of analysis, children were
categorized according to similar descriptive developmental characteristics along the three dimensions of our study: communication,
socialization, and symbolic elaboration (see Table 1). The two subgroups that emerged from analyzing the production of the images
and the story progression were children that were characterized
by (1) pre-symbolic expression and (2) symbolic expression. The
pre-symbolic group was further categorized into two sub-groups
of tactile or sensorial exploration and functional play. The symbolic group revealed three sub-categories of rigid and ritualistic
play, beginnings of storytelling, and symbolic themes organized
into story form. From each group, an analysis was made of the
common developmental capacities and symbolic themes and tendencies demonstrated in the sand tray.
General ndings
The sandplay workshops appealed to the children, who manifested their appreciation by participating with enthusiasm, joy and
pride in both the rituals and the sandplay activity. Their teachers reported that they anticipated and talked about the workshops
several days in advance.
Within the 10-week period of the workshops no negative or
regressive reactions, such as poor concentration or increased agitation in class following a session, were observed or reported by
the teachers concerning the childrens behaviour. It was possible to
note changes in the complexication of symbolic use in the sandtrays and the stories told, as well as improvement in the general
attention and participation of most children. For instance, several
highly avoidant children engaged in social play, others moved from
functional play to incorporating some symbolic representational
play, while others elaborated on their symbolic and imaginary play.
The rigid play of certain children began to demonstrate more exibility as new elements borrowed from other childrens sandtrays

Fig. 1. Session 1Sheldons line of glass beads.

59

were incorporated into their play. Most children demonstrated

some awareness of their peers through their practice of mirroring or
imitating their choice of gurines, their use of water, the way they
structured or placed gurines in their sandtrays, and through their
attentive listening to their peers stories. Their play was enhanced
as they built on each others ideas through this social interaction.
The children tended to progress, as is particular to sandplay, in a
spiral rather than a linear fashion (Weinrib, 1983/2004). In this way,
the same sandtrays can be reproduced from week to week with only
minor changes and then move to a new level of development as the
sandplayer integrates new experiences in their creative process.
Further descriptions of the ndings highlight the specic details of
the two sub-groupings of pre-symbolic and symbolic expression,
with vignettes to illustrate the ve sub-categories.
Pre-symbolic expression
Tactile and sensorial exploration: non-verbal (naming some
needs)
The pre-symbolic play of children of this rst category (n = 3)
was characterized by engaging on a sensory level with the sand
and a few objects with tactile qualities. There were some indications of functional play, such as moving a bus back and forth.
Occasionally, one child would draw simple faces in the sand; however, for all three children there was no sustained creation in the
sandtrays, nor any symbolic interactions with the gurines. Water
played an important function in attracting and sustaining interest
and concentration in the activity.
Communication, which was predominantly non-verbal, was
sustained through the medium of sandplay by exploring the properties of the dry and wet sand and of certain tactile objects by
squishing, rolling, and hiding them in the sand with the child, or
by playing directly with him by hiding and catching his hands in
and under the sand. The autonomy of each child differed, ranging
from an ability to sustain attention as the workshops progressed,
to requiring one-on-one interaction with the therapist/teacher to
stimulate joint attention. Children were rarely able to engage with
other peers for longer than a few interactions.
Engaging these children in the opening rituals proved more challenging but was facilitated through clear directions supported by
pictograms and signing made available by the teachers and special educators. The closing ritual, on the other hand, because of its
unchanging nature and group dynamic, proved largely popular and
two of the boys began actively participating in the activity as they
became more engaged and accustomed to the routine.
Vignette: Sheldon. Sheldon, a 9-year-old boy who was present for all
10 sessions, had a tentative initial contact with sandplay. He played
with some sticky, stretchy worms and was given a small quantity
of water to rinse them with. On this rst day he merely poured the
water into the sand. His attention was very limited although he
lined up some marbles a teacher offered him (Fig. 1). Sheldon was
very restless during the rst four sessions leaving his seat to look
for more water or for something else to do, making loud lamenting
sounds and dropping toys on the ground. Although Sheldons verbal
communication was very limited he did ask for water to play with
each week. Water became a means of engaging with the sand, by
changing its properties; as well as engaging him with people as he
expressed his need for it.
Apart from occasionally functionally playing with a school bus,
the majority of Sheldons play was sensorial. The therapist elaborated upon his play with marbles and bright, sticky creatures, at rst
by moving the worms along his arms, which he enjoyed. When his
teacher began playing a hide-and-seek game with him by hiding her
own hands in the sand and touching Sheldons hands beneath it, his
attention was fully aroused and an intense non-verbal interaction

60

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

Fig. 3. Session 2Ryans functional play with a bus.

Fig. 2. Session 5Sheldon engaged in sensory play.

was established. Each subsequent session began with him pouring

water in the sandtray and mixing it through burying his hands in
the sand and inviting, through eye contact, for them to be found. He
was then able to play on his own with the worms and marbles for an
extended period of time (Fig. 2). As he began to experience some
shared attention and became more regulated through the structure provided, his affect became more positive and his attention
increased to the duration of the 45-min workshop. Sheldon, who
spoke rarely, began saying bye-bye to his sandtray each time he
left, perhaps indicating an emotional investment in the activity of
sandplay.
Pre-symbolic/functional play
Each child of this sub-category (n = 4) was able to move from
pre-symbolic functional play to the beginnings of symbolic play. In
most cases, the children were limited in their verbal expression.
They communicated with a few short phrases, some responded
with echolalia, and some were able to use appropriate greetings
such as hello, how are you?. Their participation was much more
expressive during the rituals than in the less practiced context of
talking about their play. Occasionally, indications of shared meaning in communication within the play were present. All the children
in this group beneted from social interaction, mostly with teachers and therapists, but also from imitating other childrens work,
such as adding water to the sand. The childrens language became
more articulate during their play as they began to master what they
were doing.
Their initial play was characterized by a sensory exploration of
the gurines and the qualities of the sand and some functional play.
For several weeks the children would resist attempts to extend and
expand on their play, closing circles of communication very quickly.
Through close observation of the play and placing a prompt at the
right moment, eventually the point of entry would be found and
the childs play would move to a deeper level of complexity. No
sustained storyline was achieved for this group but rather pieces of
stories were created: the animal drinks from the well; The Ford
pick-up lives at this house; the bus was attacked by a crocodile.
All the boys, through much trial and error in extending and
expanding their play, reached, by the end of the 10 sessions, a level
of representational play conducive to the elaboration of a more
complex symbolic play and storyline.
Vignette: Ryan. Ryan is a timid, gentle 9-year-old boy, who had a
strong tendency to respond with echolalia. He was present for all
10 sessions. He agreeably took any gurine offered to him but soon
put it down and continued functionally playing with a school bus
or re truck, which he moved timidly around the sandtray (Fig. 3).

For over half of the sessions, Ryan politely rejected all attempts
to expand on his play and repeated back all that was said to him in a
quizzical way. He responded with mild interest when the therapist
mirrored his play and moved another bus following him in the tray.
An entry into his play happened when the therapist elaborated on
his school bus play by placing a house as a destination for the bus to
stop. Ryans attention was piqued only when a personal connection
was made as it was suggested that Ryan himself lived in the house
and would be picked up by the bus. When the therapist encouraged
him to look in the gurine bin for a suitable representation for his
school, he appropriately picked a very large house. Further elaborations stimulated his active participation, and spontaneous symbolic
representation manifested as Ryan, in a quick and excited manner,
picked out animal gures from the gurine bin and identied each
one as his classmates and special educator, placing them around
the school in his tray. His teacher mirrored his excitement as he
named the familiar people represented in his sandtray. This activity marked the beginning of Ryans spontaneous symbolic play, as
well as a signicant reduction of his echolalia during his play. In
the remaining sandtrays, Ryan began each sandplay with placing
the house, school, and bus in the sandtray. Non-prompted spontaneous symbolic play began to emerge in such scenes as the bus
being attacked by a giant crocodile, re trucks and remen being
added to the sandtray, skeletons being buried, and so on (Fig. 4).
Ryan was able to develop his play after making an initial emotional
investment that had personal resonance for him.

Fig. 4. Session 9Ryans emotional investment to the play.

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

Fig. 5. Session 2Marios rigid and ritualistic sports game.

Symbolic expression
Emerging symbolic, ritualistic and rigid play
This category of emerging symbolic play (n = 6) illustrates a
group of children that display the rigidity and ritualistic play, or
what might be termed as perseverative play (Pilewskie, 2007), often
characterized in children with autism (Wolfberg, 1996). This group
of children often had a difcult time regulating their emotions and
their level of internal organization or disorganization was often represented in the sandplay image. For example rigidity and ritualistic
play was demonstrated by nearly identical placement of gurines
with a repeated set of actions or scripts in each session that had
little variation or exibility. Disorganization was apparent through
chaotic placement or randomly lling the tray full of gurines.
In terms of their level of communication, they would often
express their interest and engagement non-verbally and were able
to say a few words or a sentence to describe their image or story
when prompted. Interactions during the sandplay were limited
to a few exchanges, as it was difcult for the teachers and facilitators to nd a point of entry into the play or ask questions to
the child, because of the single-focused quality of the play. When
the therapist was able to enter into the play, such as playing the
referee in a scripted sports game, it was possible to use the play
as a form of emotional co-regulation. However, for all the children in this category, they were often quite social and expressive
during the opening and closing ritual, demonstrated by smiling,
making eye contact, and initiating appropriate answers to our interactive themes and imaginary play. In their symbolic expressions
they would create simple storylines and symbolic themes such as
aggression in a sports game, or nurturance through an interactive
scene of gurines set amongst home furniture and food; however,
no further verbal description would be made.
Vignette: Mario. Mario is a 10-year-old boy who appeared smiling,
content, and interactive with his peers, teachers, and facilitators,
especially during the opening and closing ritual. However, from the
rst session Marios play in the sand tray was characterized by rigid
placement of gurines between two goal posts playing sports.
Throughout the play he seemed to perseverate by remaining xed
and focused on placing his team of players in the tray. He would
not respond to prompts or questions, nor was he interested in the
sandplay images of his peers and would often resist nishing his
play by the end of the session. When he did respond to our questions
of what was happening in the tray he would say They are playing
sports (Fig. 5).
Mario would ritualistically repeat this play in every session,
gradually modifying the play by adding animals, cars, play money,
and ags in the tray, and often there were so many gurines that
more and more would be placed outside of the sandtray. Mario

61

made attempts towards symbolic interaction by placing the gurines facing each other; however, his rigid play did not provide
him with the fulllment of symbolic engagement. Yet in the last
session Mario appeared to satisfy this need by uncharacteristically
asking to be placed in the spot of his friend George, with whom he
had interacted in the previous session. He imitated Georges play by
having the gurines bounce off the sides of the box like a wrestling
ring to attack each other, while also borrowing English phrases
from George, like excitingly calling out 1-2-3 and Oh my God
as the gurines would launch an attack. Mario began the sessions
playing in a ritualistic and rigid manner; however, he responded
to his desire for greater symbolic engagement through imitation
and borrowing another play script in order to move beyond his
perseveration in his play. Despite the repetition and rigidity of
his sandplay images, Mario demonstrated an untapped capacity
for symbolic expression and engagement in the play and would
have probably beneted from a longer series of sessions in order to
elaborate his play.
Symbolic representation and the beginning of storytelling
In this group of children (n = 7), the sandplay images are richer
in symbolic representation and through the process the children
began to demonstrate a better capacity to link elements into a
rudimentary storyline. They had varying degrees and capacities of
communication, some were able to verbally describe their sandplay
images in full or partial sentences with and without prompting,
while others would enact or play out their stories in their telling.
For the most part, the children in this group were able to interact
during the exercises and enjoyed telling and showing their sandplay stories to their peers at the end of the session. Three of the
children actively played in each others sandtrays and told stories together. Many of them would incorporate gurines or stories
that would be inspired from the exercises or from observing other
childrens story images.
The symbolic themes in the sandplay of this group ranged
in complexity and included age-appropriate themes of aggression/ghting/death/fear; accidents and rescue stories; sharing and
nurturance; repairing when characters were hurt; and activities
from their daily livesschool, eating, sports, and birthdays. In this
group there was a range of different intellectual abilities, which
reected the degree of complexity that children were able to produce in their symbolic play. The story telling was not always
coherently organized into a recognizable story format with a beginning, middle (conict/issue), and end or resolution, but the stories
would often closely approximate this format. This group of children would be able to respond to play invitations from their peers,
teachers, or facilitators with common shared meanings and to elaborate upon them to continue the play. Through the progress of each
session, it seemed that the majority of children were able to elaborate their play by adding more elements and investment in their
stories.
Vignette: Emile. Emile is an 11-year-old boy, who was timid, calm,
and attentive throughout the sessions, and who liked to engage in
the exercises and activities. He communicated, often with prompting, in short, concrete phrases, seemingly self-stimulating on words
at the end of his sentences that he would accentuate with an inection or repetition. In his rst and second session, he demonstrated
elaborated functional play when he lined up the houses around
the sandtray and repeatedly moved a school bus back and forth
on the L-shaped corridor he created (Fig. 6). He allowed the therapist to elaborate the game by adding a trafc light, but when asked
about his story, he said The school bus is making his rounds in the
neighborhood houses and would not elaborate further.
However, by the third week, a shift in Emile became evident as
he was inuenced by the imaginary play theme of animals in our

62

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

Fig. 6. Session 1Emiles school bus in the neighborhood.

opening ritual. He initially began playing with the school bus, but
soon became more invested in creating a forest scene of animals
and plants. Initially, he was very timid to tell his story, but with
prompting from his peers he proudly described his story image stating The sun and stars are looking at the springtime. There are birds
and caterpillars in the forest. There are birds in nests in the forest.
There is a y buzzing (making the sound). Although he repeated this
springtime image for another two sessions, it appeared that Emile
became more inuenced by the imaginary play in the rituals and
slowly began incorporating people and creating elaborated scenes
(Fig. 7). As his stories developed, Emile gained the condence and
desire to recount his stories, often going beyond the time limit available. By the end of the ten sessions, Emile was able to elaborate his
symbolic play by incorporating more interactive elements, while
also beginning to link them coherently into a storyline.
Symbolic themes organized into story form
This last category regroups children (n = 5), who demonstrated
the capacity to create spontaneous, complex scenarios involving
much detail, drama, and a mostly coherent storyline. Emotional,
concrete, and magical thinking were present in their sandplay.
Many themes were developed expressing loss, intrusions, competition, conict resolution, and the establishment of boundaries.
This group had a higher level of language mastery, and
expressed relatively coherent stories in complete sentences.
Although they had the expressive vocabulary to indicate emotions, such as fear or sadness, affect was rarely present in their
expressive tone.
As demonstrated in sandplay with children within other classroom settings (Lacroix, 2002), three children appeared to use
sandplay to work through personal struggles, such as difculties
at home, frustrations with physical limitations, or personal tension
with classmates.

Fig. 7. Session 5Emiles springtime and people.

Fig. 8. Session 3Isaacs detective story.

Children sought interaction with the therapists and teachers,

and enjoyed communicating their stories to the class. Children
of the 4th and 5th category were very appreciative audiences for
their peers: they laughed, asked questions, expressed admiration
as well as compassion and sadness. In a very few cases, irritation
or mockery was expressed as well. Children of this group participated actively in rituals often offering personal anecdotes, creating
collective stories, and participating in pretend play such as hiking through a mountain and camping in a cave. Themes from the
opening ritual were often present in the sandplay.
In summary, children of this sub-group were able to express
complex emotional content and link several ideas into a relatively
coherent storyline. The sandplay offered them a medium to elaborate on this capacity, in some cases write stories from their sandplay
(often for the rst time), as well as providing them a medium to
interact with peers and teachers.
Vignette: Isaac and Jean-Claude. Isaac is a thoughtful, articulate 11year-old boy who often verbalized his actions of play with private
speech. Like many of his classmates, he initially enacted a game
of soccer. By the third session, Isaac stopped himself midway setting up another game saying, no, I want to tell a story, choosing
to let go of the familiar sports script. He then created a dramatic
story involving kidnapping, detective work, and supernatural interventions, and he connected with the emotions of his characters
relating, He did not nd his grandmother and began to cry (Fig. 8).
Role-playing and pretend imaginings involving object substitution and symbolic thinking were demonstrated in subsequent
sandplays created with his fellow classmate Jean-Claude, a 12-yearold boy with a keen sense of humor who often initiated invitations
to play with Isaac. Isaac solved the problem of having Jean-Claude
idly wait while he narrated the story, by giving his classmate the
role of cameraman. Isaac turned a mirror upside down and the hole
in the handle became the lens through which either child could lm
the others story. Jean-Claude, although feeling a bit self-conscious
at rst, lmed Isaacs story and, in this manner, his symbolic play
was enhanced (Fig. 9). Isaac, through this game, began to express
a meta-cognition of play in the sense that he was able to step out
of his play, observe his play and attempt to share his symbolic play
with his peer.
The boys complemented each other as Isaac contributed elaborate storylines that tended to get bogged down in detail whereas
Jean-Claude added drama and realism through such devices as robberies and assigning more logical outcomes to the stories, such as
the police arriving to assist the victims. Some of the themes of
the stories sparked interesting discussion about their knowledge

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

Fig. 9. Session 8Isaac and Jean-Claudes shared story and lm.

of religious practices or preferred movies. The medium of sandplay allowed these higher-functioning children with autism a space
to interact and focus joint attention on a creative activity while
supporting and stimulating the symbolic interactions in their play.
Discussion
As pleasure and enjoyment are essential characteristics of play
(Wolfberg, 1996), it is noteworthy that the children with ASD who
participated in this program enjoyed the sandplay process. Indeed,
sandplay was originally devised as a safe, free, and unfocused
space for children to play (Mitchell & Friedman, 1994). As such,
sandplay, adapted as a semi-structured creative intervention for
children with ASD, provided a multi-layered support for play and
creative expression. The sand and the diverse array of gurines
supported sensory and symbolic play; the individual sandtrays
delimited a protected space; while social interaction and expressive
communication was encouraged through the sharing of gurines,
the imaginary play in the rituals, and the storytelling exchanges.
Within the structure and framework of the rituals and sandplay,
children were given a non-goal-oriented and creative space for
unstructured play that encouraged symbolic development at their
own rhythm. This is an important condition for supporting creativity that is absent in adult directed activities typically practiced
within programs for children with ASD in schools. This complementary space to the regular academic program allowed freedom
for children to return to sandplay images and themes on a weekly
basis without the expectation to achieve a set academic goal.
Children responded to this semi-structured sandplay activity
with initial tentative involvement, yet over the course of the 10
weeks the increased engagement and investment in the activity
supported their developmental skills in communication, socialization, and symbolic elaboration. Children tended to work in a spiral
rather than a linear fashion, staying with the same themes and
building and expanding, with growing exibility upon their play
capacities over the course of the program. Children could address
their particular limitations pertaining to the foundations of relating, communicating, and thinking (Greenspan & Wieder, 2006) that
had not been previously mastered, such as sensory exploration,
engaging and relating, sharing attention, and purposeful emotional
interaction.
It appeared that for the children in the pre-symbolic level of
expression, the sensory tactile aspect of the sandplay encouraged motor skills development and early social play, while also
encouraging emotional regulation. Therapists, teachers, and special educators played the role of supporting emotional regulation as
well as two-way engagement by enhancing and expanding on sen-

63

sory exploration and functional play through making an emotional

connection to the childs experience. According to developmental
theories of play (Greenspan & Wieder, 2006; Wolfberg, 1996), a
childs ability to engage in higher levels of spontaneous communication, socialization, and symbolic elaboration is based on shared
attention and sustaining two-way pre-symbolic communication.
The sandplay intervention demonstrated that it could be a medium
used to develop the core decits of children with autism, while at
the same time supporting children operating at higher functional
emotional levels within the same classroom setting.
Children expressing at a symbolic level of play and story development appeared to benet from the sandplay intervention by
having the visual gurines and objects to sustain the development
of their emotional thinking and linking abstract ideas into coherent storylines. As children became familiar with the medium and
gained condence in creating sandplay images, they became more
at ease with interacting with each other, listening to each other, and
being able to narrate their stories for their classmates and teachers.
The children that had the least progression in this category were
children showing rigid, ritualistic, and perseverating tendencies in
their play, which made it challenging for therapists to intervene and
nd a point of entry into their play. However, just as the other children in this category who functioned at higher levels of symbolic
expression beneted from the shared attention of the sandplay and
social interaction from peers and adults during the intervention,
it seems promising that this group of children could also benet from more sandplay sessions to develop theses skills of play.
More research is necessary to investigate how sandplay could provide the medium and social interaction to interrupt perseverative
play; however, it seems to support current theories (Sherratt, 2002;
Wolfberg & Schuler, 2006), that symbolic play can be enhanced by
watching or interacting with peer models.
Sandplay was adapted as a group creative play support program
using a relational approach to play (Greenspan & Wieder, 2006) in
a semi-structured format, in order to provide the structure that
is the cornerstone of programs for children with ASD (National
Research Council, 2001), while also providing an open-ended, creative space to support spontaneous communication, socialization,
and symbolic play. Sandplay offered a unique combination of sensory stimuli (sand, water, colorful, and textured objects), the raw
material for story making (gurines, building materials), containment and a free creative space, as well as social contact and
stimulation. In this sense the structure and materials supported
the relational approach elaborated in Greenspans (2006) Floortime
method, but structured and supported the ability to work with a
small group of children with varying degrees of developmental abilities. Sandplay adds a relational triangular component, in the sense
that the interaction is through the sandplay, whereas Floortime is
designed as an interpersonal and open-ended dynamic.
The use of an action-research methodology provided the framework that allowed for teachers input into the program and
adjustment by the therapists to the various developmental needs
of the children. Teachers were invaluable contributors as their
knowledge of the children helped to identify types of play that
they enjoyed, such as Sheldons hide-and-seek play, and provided
insight into how the childrens play reected their interests or personal experiences.
The teachers introduction of a story structure around the play
provided positive support, for certain students, for organizing the
narration of their sandplay. Thus, teachers goal of stimulating
creative writing harmonized well with the overall objective of
stimulating symbolic elaboration and communication. However, it
appeared also that for some children this extra structure created
pressure to conform to a certain level of storytelling they had not
yet attained. The use of a minute timer provided by another teacher
to delimit the storytelling appeared to meet the childrens need

64

L. Lu et al. / The Arts in Psychotherapy 37 (2010) 5664

for structure without too much constraint or expectation. Further

investigation is needed to evaluate the impact of restraining the
number of gurines available for certain children with ASD, specifically those with the tendency to perseverate. In some cases, like
the practice of limiting the amount of water provided, this would
appear to be recommended, in other instances providing regulation
through interaction with the children through the play (providing
trafc lights, police, or referee gures) would seem sufcient, and
in the case of Mario, he appeared to interrupt his own perseverating play by imitating another childs play that he found more
stimulating.
Further research is needed to replicate this work and to conrm
the benets of providing, through such a program, a more creative,
less structured, and non-goal oriented space to meet existing educational goals for children with ASD. Once the benets of sandplay
to support creative symbolic thinking in children with ASD has
been investigated, further research is needed to evaluate to what
extent the program can be transferred to teachers to implement
and how much training, external support, and supervision is necessary to successfully disseminate the program in schools. Finally
further study is encouraged to see how sandplay adapted as a creative group activity within the classroom can be applied also to
the needs of other children who may experience other social or
emotional difculties.
In conclusion, sandplay adapted to children with ASD appears
to provide the potential of enough structure around a free and
unfocused creative space to support the developmental skills of
communication, socialization, and symbolic elaboration. Children
respond and interact with the medium at their individual developmental level, which makes it an ideal medium to be used in a school
setting for classrooms with various developmental needs.
Acknowledgements
The research ndings from this school intervention were based
on an action research investigated by school psychologist Dogratias Bagilishya, professor and art therapist Louise Lacroix, from
Concordia University, along with the Youth Mental Health Team
(CSSS), and was funded by the Quebec Ministry of Education, Sports
and Leisure. We would also like to thank the students, parents,
teachers, special needs educators and staff at cole SimoneDesjardins for their participation and support in this research.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.aip.2009.09.003.
References
Boucher, J. (1999). Editorial: Interventions with children with autism - methods
based on play. Child Language Teaching & Therapy, 15(1), 15.
Brock, S. E., Jimerson, S. R., & Hansen, R. L. (2006). Identifying, assessing and treating
autism at school. New York: Springer.
Bryson, S. E., Rogers, S. J., & Fombonne, E. (2003). Autism spectrum disorders: Early
detection, intervention, education, and psychopharmacological management.
Canadian Journal of Psychiatry, 48(8), 506516.
Carey, L. J. (1990). Sandplay therapy with a troubled child. The Arts in Psychotherapy,
17, 197209.
De Domenico, G. S. (1999). Group sandtray-worldplay: New dimension in sandplay
therapy. In D. S. Sweeney, & L. E. Homeyer (Eds.), Group play therapy: How it
works, whom its best for (pp. 215233). San Francisco: Jossey-Bass Publishers.
Friedman, H. S., & Mitchell, R. R. (Eds.). (2008). Supervision of sandplay therapy. London and New York: Routledge.
Greenspan, S. L., & Wieder, S. (2006). Engaging autism: Using the oortime approach
to help children relate, communicate, and think. Cambridge, MA: Da Capo Press.

Greenspan, S. L., & Wieder, S. (2007). The developmental individual difference,

relationship-based (DIR/Floortime) model approach to autism spectrum disorders. In E. Hollander, & E. Anagnostou (Eds.), Clinical manual for the treatment of
autism (pp. 179209). Washington, DC: American Psychiatric Publishing.
Hess, K. L., Hein, L. J., Morrier, M. J., & Michelle, I. L. (2008). Autism treatment survey:
Services received by children with autism spectrum disorders in public school
classrooms. Journal of Autism Development Disorders, 38, 961971.
Jarrold, C. (2003). A review of research into pretend play in autism. Autism, 7(4),
379390.
Jarrold, C., Boucher, J., & Smith, P. (1996). Generativity decits in pretend play in
autism. British Journal of Developmental Psychology, 14, 275300.
Kalff, D. (2003). Sandplay: A psychotherapeutic approach to the psyche. USA: Temenos
Press. Original work published in 1980
Kestly, T. (2001). Group sandplay in elementary schools. In A. A. Drewes, L. J. Carey,
& C. E. Schaefer (Eds.), School-based play therapy (pp. 329349). New York: John
Wiley & Sons, Inc.
Knoblauch, P. J. (2001). Play therapy in a special education preschool. In A. A. Drewes,
L. J. Carey, & C. E. Schaefer (Eds.), School-based play therapy (pp. 81101). New
York: John Wiley & Sons, Inc.
Lacroix, L. (2002). Retour au pays dorigine: crativit sensorielle par lutilisation du
jeu de sable en art thrapie. Prisme, 37, 3245.
Lacroix, L., Rousseau, C., Gauthier, M.-F., Singh, A., Gigure, N., & Lemzoudi, Y. (2007).
Immigrant and refugee preschoolers sandplay representations of the tsunami.
The Arts in Psychotherapy, 34, 99113.
Lan, G. (2008). Supervision of sandplay therapy in preschool education in China. In
H. S. Friedman, & R. R. Mitchell (Eds.), Supervision of sandplay therapy. London
and New York: Routledge.
Libby, S., Powel, S., Messer, D., & Jordan, R. (1998). Spontaneous play in children
with autism: A reappraisal. Journal of Autism and Developmental Disorders, 28(6),
487497.
Lowenfeld, M. (1979). The world technique. London: Allen and Unwin.
Martin, D. M. (2001). Preschool. In A. A. Drewes, L. J. Carey, & C. E. Schaefer (Eds.),
School-based play therapy (pp. 163174). New York: John Wiley & Sons, Inc.
Mitchell, R. R., & Friedman, H. S. (1994). Sandplay: Past, present, and future. London:
Routledge Press.
National Research Council. (2001). Educating children with autism. Committee on
Educational Interventions for Children with Autism, Division of Behavioral and
Social Sciences and Education. Washington, DC: National Academy Press.
Pabon, A. J. (2001). Sandplay therapy in a time-limited school-based program. In
A. A. Drewes, L. J. Carey, & C. E. Schaefer (Eds.), School-based play therapy (pp.
123138). New York: John Wiley & Sons, Inc.
Pilewskie, A. (2007). Perseveration. In B. S. Myles, T. C. Swanson, J. Holverstott,
& M. M. Duncan (Eds.), Autism spectrum disorders: A handbook for parents and
professionals (p. 245). Westport Connecticut: Praeger.
Rousseau, C., Lacroix, L., Singh, A., Gauthier, M.-F., & Benoit, M. (2005). Creative expression workshops in school: Prevention programs for immigrant and
refugee children. The Canadian Child and Adolescent Psychiatry Review, 14(3),
8285.
Sherratt, D. (1999). The importance of play. Good Autism Practice, 1(2), 2331.
Sherratt, D. (2002). Developing pretend play in children with autism: A case study.
Autism, 6(2), 169179.
Sherrat, D., & Donald, G. (2004). Connectedness: Developing a shared construction of
affect and cognition in children with autism. British Journal of Special Education,
31(1), 1015.
Stahmer, A. C., Collings, N. M., & Palinkas, L. A. (2005). Early intervention practices for
children with autism: Descriptions from community providers. Focus on Autism
and Other Developmental Disabilities, 20(2), 6679.
Steinhardt, L. (1998). Sand, water, and universal form in sandplay and art therapy.
Art Therapy, 15(2), 252260.
Stringer, E. T. (2007). Action research: A handbook for practitioners (3rd ed.). Thousand
Oaks, CA: Sage.
Tanguay, D., dAminico, M., Dolce, S., & Snow, S. (2004). Lutilisation du jeu de sable
avec une clientle prsentant une dcience intllectuelle. Revue francophone
de la dcience intllectuelle, 15(1), 2340.
Van Dyk, A., & Wiedis, D. (2001). Sandplay and assessment techniques with
preschool age children. In A. A. Drewes, L. J. Carey, & C. E. Schaefer (Eds.), Schoolbased play therapy (pp. 1637). New York: John Wiley & Sons, Inc.
Weinrib, E. L. (2004). Images of the self: The sandplay therapy process. USA: Temenos
Press. Original work published in 1983.
Wieder, S., & Greenspan, S. L. (2003). Climbing the symbolic ladder in the DIR model
through oortime/interactive play. Autism, 7(4), 425435.
Wing, L., & Gould, J. (1979). Severe impairments of social interaction and associated
abnormalities in children: Epidemiology and classication. Journal of Autism and
Developmental Disorders, 9(1130).
Wolfberg, P. J. (1996). Play and imagination in children with autism. New York: Teachers College Press.
Wolfberg, P. J., & Schuler, A. (2006). Promoting reciprocity and symbolic representation in children with autism spectrum disorders: Designing quality peer
play interventions. In T. Charman, & W. Stone (Eds.), Social and communication development in autism spectrum disorders: Early identication, diagnosis, and
intervention. New York: The Guilford Press.