The Impact of Pretend Play On Children's Development: A Review of The Evidence
The Impact of Pretend Play On Children's Development: A Review of The Evidence
The Impact of Pretend Play On Children's Development: A Review of The Evidence
Ptetend play has been claimed to be cmcial to childten's healthy development. Hete we examine
evidence fot this position vetsus 2 altetnatives: Ptetend play is 1 of many toutes to positive developments
(equifinality), and ptetend play is an epiphenomenon of othet factots that drive development. Evidence
from several domains is consideted. Fot language, narrative, and emodon tegulation, the teseatch
conducted to date is consistent with all 3 positions but insufficient to dtaw conclusions. Fot executive
function and social skills, existing teseatch leans against the crucial causal position but is insufficient to
differentiate the othet 2. Fot teasoning, equifinality is definitely supported, niling out a ctucially causal
posidon but still leaving open the possibility that ptetend play is epiphenomenal. Fot ptoblem solving,
thete is no compelling evidence that ptetend play helps ot is even a cottelate. Fot creativity, intelligence,
consetvation, and theoty of mind, inconsistent cottelational tesults from sound studies and nonteplication
with masked expetimentets ate ptoblematic fot a causal position, and some good studies favot an
epiphenomenon position in which child, adult, and envitonment chatactetistlcs that go along with play
ate the true causal agents. We end by consideting epiphenomenalism more deeply and discussing
implications fot pteschool settings and futthet teseatch in this domain. Out take-away message is that
existing evidence does not suppott strong causal claims about the unique impottance of ptetend play for
development and that much mote and bettet teseatch is essential fot clatifying its possible tole.
How does pretend play affect children's development? Claims development . . . so important . . . that it has been recognized by
for its positive impaet are resounding. The National Association the United Nations High Commission for Human Rights as a right
for the Edueation of Young Children, the major presehool accred- of every ehild"^ (Ginsburg, the Committee on Communieations, &
iting body in the United States, stated in its recent position paper, the Committee on Psyehosoeial Aspects of Child and Family
"high-level dramatic play produces documented eognitive, social, Health, 2007, p. 182). P. K. Smith (2010, pp. 28-29) gave many
and emotional benefits" (Copple & Bredekamp, 2009, p. 15). An other examples of the important and wide-reaehing benefits attrib-
article aimed at parents states that play "is a significant contributor uted to pretend play (see also Bredekamp, 2004; Brown &
to the child's eognitive, physical, emotional, and social develop- Vaughan, 2009; Elkind, 2007; Tullis, 2011). American parents
ment"' (Hurwitz, 2002, p. 101). Some even maintain that pretend eoneur (Roopnarine, 2011), and ehild development experts en-
play's impaet is unique: A elinieal report on the subjeet for the dorse pretend play even more strongly (K. R. Fisher, Hirsh-Pasek,
American Association of Pediatrics opened, "play is essential to Golinkoff, & Gryfe, 2008). Entire preschool curricula are designed
around pretend play because of the "unequivocal evidence for [its]
eritieal importance" to children's development (Zigler & Bishop-
This article was published Online Fitst August 20, 2012. Josef, 2004, p. 9). Master teachers' diseussions of why pretend
Angeline S. Lillatd, Matthew D. Letnet, Emily J. Hopkins, Rebecca A. play is so vital for ehildren are convincing (e.g., Paley, 2005), and
Dote, Etic D. Smith, and Carolyn M. Palmquist, Department of Psychol- we agree: When we wateh ehildren in pretend play, it seems to us
ogy, University of Vitginia. like a very important activity.
Matthew D. Letnet is now at the Department of Psychiatry and Behav-
However, many non-Anglo eultures do not share this view of
iotal Neutoscience, Univetsity of Chicago Medical Centet, Chicago, E..
Preparation of this article was supported by National Science Founda- pretend play's importance, and perhaps as a result, ehildren grow-
tion (NSF) Grant 1024293, a Btady Education Foundation gtant, and a ing up in those eultures pretend mueh less (Gaskins & Goncu,
University of Vitginia Sesqui award to Angeline S. Lillatd; the Ametican 1992; Laney, 2007). A recent survey found that in only five of 16
Psychological Foundation's Elizabeth Munstetbetg Koppitz Fellowship,
the James H. and Elizabeth W. Wtight Endowed Fellowship from the
Jeffetson Scholars Foundation, and gtants from the Ametican Psycholog- ' Elsewhete it is cleat that ptetend play is intended; fot example, "in
ical Association and the Association fot Psychological Science to Matthew play, everything and anything can happen: a sheet ovet a table becomes a
D. Letnet; and an NSF graduate fellowship to Etic D. Smith. castle" (Hutwitz, 2002, p. 101).
Cottespondence concetning this article should be addtessed to Angeline ^ Ptetend play is Ginsbutg et al.'s (2007) focus; for example, "play
S. Lillatd, Department of Psychology, Univetsity of Vitginia, P.O. Box allows children to create and explore a world they can master, conqueting
400400, Chatlottesville, VA 22904. E-mail: [email protected] their fears while practicing adult roles" (p. 183).
LILLARD ET AL.
countdes surveyed (the United States, the United Kingdom, Ire- it does continue into middle childhood and beyond (E. D. Smith &
land, Portugal, and Argentina) do the majodty of mothers say their Lillatd, in press).
children (ages 1-12) often participate in imaginative play (D. G. There are several other forms of play besides pretend (see
Singer, Singer, D'Agostino, & Delong, 2009). Even within the Pellegdni, 2009; P. K. Smith, 2010); in particular thete is a small
United States, thete is individual vadation in how much children but impottant litetatute on physical play (such as hopscotch and
pretend (Fein, 1981). Should infrequent pretenders be pretending tough-and-tumble play), which has been well teviewed elsewhere
mote? Would doing so help their development? Is the evidence (Pellegdni & Bohn, 2005; Pellis & Pellis, 2009). Such forms of
strong enough to warrant designing curdcula around pretend play play assist sustained attention in conventional school situations
and dedding preschools that do not encourage it? Here we examine (Pellegdni & Bohn, 2005); they also (in the case of tough-and-
evidence cited in support of pretend play's importance to deter- tumble play fighting) assist emotion regulation, social cootdina-
mine whethet thete is a convincing case. The evidence concems tion, and normal sexual behavior, at least in some rodents and
six domains of development, chosen because they ate frequently nonhuman pdmates (Pellis & Pellis, 2009). Pretend play can
claimed to be assisted by pretend play (e.g., see Ashiabi, 2007; overlap with these and other types of play. For example, physical
Bergen, 2002; Ginsburg et al., 2007; Hirsh-Pasek, Golinkoff, Berk, play overlaps with pretend play when children pretend to be
& Singer, 2009; Isenbetg & Quisenbetry, 1988; Lillard, 2001a) fighting wardots. Object play overlaps with pretending when a
and because we found at least a half dozen studies conceming child animates those objects.
each: nonsocial cognitive aptitudes (with five subdomains), social The literature is not always cleat as to when pretend play
cognition, social skills, language, narrative skills, and self- specifically, versus play mote generally, or some other specific
tegulation (with the subdomains of executive function and emotion type of play is at issue; this can be seen in the quotes with which
regulation). we opened (but see footnotes 1-2), and probably adses because
First we define pretend play and review three theoretical posi- young children's play is so often infused with pretense. Out aim
tions on whether and how it affects development generally. Next, here is not to resolve this ambiguity but rather to considet studies
we descdbe pattems of evidence that would suppott each position, used to support claims that play is crucial to positive develop-
then review and discuss the evidence domain by domain. Finally ments, excluding the physical play literature just mentioned, and
we consider one position more deeply, address the implications of retaining focus on pretend play as much as possible. Our main
our review for educational settings, and make suggestions for exception to this is in two subdomains of nonsocial cognitive
future research on this topic. aptitudes, creativity and problem solving, because for those skills
several studies concerning manipulative play with small objects
(which migbt or might not involve pretending) ate often cited as
Defíning Pretend Play showing play's cognitive benefits. When a study contrasted pre-
tend play with some other form of play (like construction play, as
A preliminary issue is to define pretend play. Play itself is a in building with blocks) we focused on the ptetend condition.
notodously difficult concept to pin down (Burghatdt, 2011). For Many studies strain the voluntary aspect of play in that children
our purposes the four cdteda of Krasnor and Pepler (1980) will were told to play or were instructed in acting out a story, but
define play: flexibility, positive affect, nonliterality, and intdnsic because those studies have been cited as showing play's benefits,
motivation (cf. Sutton-Smith & Kelly-Byme, 1984). Flexibility they ate reviewed here.
denotes that play behaviors vary from real ones in form (they To locate studies, the first author began with references sup-
might be exaggerated, or truncated) and/or content (one might play porting claims of play's benefits in articles like those in the
at eating with a stick instead of a spoon). Positive affect touches on opening paragraph, then back tefetenced those studies in a snow-
the idea that people look like they are having fun when they play. ball fashion. Through this process she atdved at the six main topics
Nonliterality refers to the fact that, in play, behaviors lack their and six subdomains of nonsocial cognitive skills; the subdomain of
usual meaning while paradoxically retaining it; Bateson (1972) mathematics was subsequently eliminated due to an insufficient
famously pointed out that, "the playful nip denotes the bite, but it number of studies. Ftom there a search engine (Google Scholar)
does not denote what would be denoted by the bite" (p. 317). was employed, searching by keywords ("social skills, ptetend
Intrinsic motivation suggests voluntadness: One engages in the play") and the "tefetenced by" and "related articles" features, as
activity by choice for its own sake. well as continuing to back reference ftom within articles. To avoid
Pretend play activities are the subset of play activities chatac- an unwieldy teview, we passed over studies of atypical populations
tedzed by an "as-if ' stance (Garvey, 1990). Beyond being simply or cultural vadation, and largely confined ourselves to published or
nonliteral, in pretend play a "pretense" is layered over reality in press peet-reviewed studies.''
(Austin, 1979); specifically, a pretender knowingly and intention-
ally projects some mentally tepresented altemative on to the pres-
ent situation in the spidt of play (Lillatd, 1993). Sometimes Theoretical Background
pretend play is social: A group of children share an alternative
P. K. Smith (2010) laid out three theoretically possible relation-
reality that they project, perhaps acting like they ate different
ships between ptetend play and positive developmental outcomes.
people in another place and time. Other times pretending is a solo
activity. Pretend play can involve projecting imaginary objects and
properties, or using one object as if it were another (Leslie, 1987). ^ Exceptions were made for particularly important unpublished studies
It is most prominent in eatly childhood, with ages 3 to 5 being reported in other published work by the author or thesis advisor, and for
declared its "high season" (D. G. Singer & Singer, 1992), although one article under review.
PLAY DEVELOPMENT
The first is that pretend play is crucial to optimal development. fails to evince a significant effect. This might occur, for example,
The second, which Smith supported, is equifinality: Pretending when there is substantial multicollineadty, or when an alternate
helps some developments, but it is only one possible route. Other predictor's effect is much larger, masking pretend play's effect.
activities can work as well or better. The third possibility is that Thus equifinality does not insist on 100% consistent results, but it
pretending is an epiphenomenon or byproduct of some other generally expects them.
selected-for capability, but in and of itself makes no contribution The third, or epiphenomenon, position is supported if pretend
to development; rather, the other activity or condition to which it play coincides with some other causal circumstance; in such cases
is sometimes attached is the actual contributor. Two major devel- pretend play might mistakenly be considered causal. For example,
opmental theorists, Vygotsky and Piaget, align with the first and if social pretend play is related to theory of mind because adults
third views, respectively. who engage in a lot of mental state talk also happen to encourage
For Vygotsky (1978), pretense has a crucial developmental role, pretend play, then perhaps what is actually leading to the increased
because it is the activity by which children learn to separate theory of mind is not the pretend play, but the mental state talk;
referent from object. In play, children first understand that actions social pretend play is secondary or epiphenomenal to the mental
(and objects on which one might act) can be separated from reality state talk-theory of mind relationship. If pretend play is an epiphe-
and can be based on the meaning of a given situation rather than nomenon then one might find inconsistent correlations with outcomes
on the physical properties of objects (Vygotsky, 1967). In this way, (because pretend play does not always go along with the real predic-
for example, a banana could become a phone in a pretense situa- tors) but consistent correlations between real predictors and outcomes.
tion and the child could act on it as if it were a phone, inhibiting Because different studies measure different possible predictors, the
how he or she would act on it if it were a banana. The upshot of true predictors might not always be evident. Here we evaluate the
this is that children develop abstract thought through pretend play patterns of evidence with an eye to each of these positions. Before
(Vygotsky, 1967). In addition, because reality must be inhibited, beginning to do so, it is useful to note some recurring problems in this
children also develop inhibitory control through pretending literature (see also Cheyne, 1982).
(Bodrova & Leong, 1996). Because of these features, "in play, it is
as though [the child] were a head taller than himself (Vygotsky,
1978, p. 102); play takes a child to the upper end of his or her Common IVIethodological Problems
"zone of proximal development" (p. 86).
Several problems recur in the literature on whether pretend play
In contrast, for Piaget (1962), pretend play is more an index than helps development. Sometimes these problems occur because the
a promoter of development.'' Its appearance around 18 months
indicates the development of the semiotic function, which also
allows for deferred imitation and language. The semiotic function '' Others have claimed Piaget gives pretense a stronger role in develop-
separates an idea from its referent, a memory from its context, and ment; for example. Singer and colleagues, citing Piaget (1962), claimed he
an object from its label, allowing one to entertain and elaborate on "concluded that play was a vital component to children's normal intellec-
mental content that is separate from the physical, present reality. tual and social development" (D. G. Singer et al., 2009, p. 285). In our
reading the closest Piaget (1962) comes to this is when he says it is
Here we consider which of these views is best supported by the
undoubtedly "a preparation for imaginative aptitudes" (p. 155), where
evidence. Each view is compatible with a particular pattern of imagination (as in pretend play) is the assimilative pole of thought (in
evidence from correlational and experimental (short-term and contrast to accommodation), and creative imagination arises only when one
training) studies, shown in Table 1 (cf. P. K. Smith, 2010, Table integrates the two. This is essentially the position taken by Harris (2000)
9.2, p. 187). These patterns assume methodologically sound stud- and D. G. Singer and Singer (1992): Pretending assists imagination. But
ies including sufficient duration and sample sizes. First, if pretend whereas for these modem authors this is a reason to centralize pretend play,
play does crucially cause positive developments (Vygotsky's po- our reading of Piaget's text on play suggests that this role in imaginative
sition), then strong positive correlations between pretend play and development was a minor concem; pretend play was primarily an offshoot
those developments should consistently be found; if a child pre- of the symbolic function. Perhaps confusion has arisen because elsewhere
tends more, whether naturally (in a correlational study) or due to Piaget assigns manipulative activities (Piaget, 1929) and peer interaction
(Piaget, 1932) as important to development, and pretend play often in-
an experimental manipulation, the development should increase. If
volves these other activities. But in considering manipulative activity,
pretend play causes creativity, then children who pretend more will Piaget refers more to what is now referred to as embodied cognition
generally be more creative. Additional predictors, like intelligence, ("manual work is essential to the child's mental development"; Piaget,
are also possible, but if pretending is truly the important causal 1929, p. 383), and regarding peer involvement Piaget's own focus on
factor, the unique and important relationship to pretend play pretend play was particularly as a solitary activity. Piaget (1962) did think
should hold even when those other predictors are partialled out. pretending served an egoistic function in that it allowed the child to fulfill
wishes that he or she could not fulfill in reality. A child who wants to be
Conversely, if Smith's equifinality position is correct, then one
a mother can simply pretend to be one. But Piaget was concemed with
would generally expect positive relationships between play and the
cognitive development, not personality development, and pretend play was
outcome but also correlations with other predictors that engender pre-operational because it indicated what the child lacked. For Piaget,
the outcome. For example, if social pretend play develops theory children outgrow pretending as they develop the ability to accommodate
of mind and so does adult talk about mental states, then correla- reality. Here he followed some major figures of his time in child psychol-
tions should be found for both variables. Interventions increasing ogy, such as Freud (1955, as discussed by Harris, 2(X)0) and Montessori
mental state talk and pretend play might have an additive effect (1989). Aligning with our own reading, Sutton-Smith (1966) colorfully
when combined, which could lead to even larger effects (but not if summarized Piaget's view of pretending as "a buttress to an inadequate
there was a ceiling on development for that age). There could be intelligence" (p. 108). For further discussion, see P. K. Smith (2010, pp.
cases when although equifmality is the best model, pretend play 31-37).
LILLARD ET AL.
Table 1
Three Possibie Reiationships Between Pretend Piay and Deveiopment
Expected pattern of results Crucial (Vygotsky) Equifmality (P. K. Smith) Epiphenomenon (Piaget)
Correlational studies Strong, unique, and consistent. Generally consistent but not unique. Inconsistent, but consistent with other variables
Including other causal variables that are causal. For example, if presence of
could mask pretend play's certain toys increases pretending in children
effects, so correlations could be who are more creative, but other objects
inconsistent. have no impact, then correlations between
pretending and creativity will be seen only
in environments with those toys.
Experimental (short-term and Strong, unique, and consistent. Strong and consistent but not Effects found only if the crucial underlying
training) studies unique, so other conditions could factor(s) is (are) influenced by the
also affect development. For intervention. For example, suppose pretend
example, skills training and play only assists development when
pretend play training could both intensive adult interaction is part of that
increase the development. training; when children pretend but there is
no intensive adult interaction, the pretend
play does not increase the development; in
addition, another condition might show that
intensive adult interaction alone increases
the development, even in the absence of
play.
research was conducted when expedmental standards were not as of experimenters being "masked" insofar as possible: that is,
high as they are today, pointing to the need to modernize the unaware of (a) the hypotheses being tested and (b) participants'
evidence base. In more recent studies, perhaps scholars did not conditions. Yet cognitive development research rarely uses
apply more dgor because of a deep belief in the power of piay masked experimenters. This might usually be fine: Child de-
(Elkind, 2007), what P. K. Smith (1988) dubbed the "play ethos" velopment researchers and the kinds of tests they give might not
and Sutton-Smith (1995, p. 279) the rhetodc of "play as progress." be vulnerable to experimenter bias under the usual circum-
Here we stdve to overcome the tendency to favor pretend play by stances. For example, we know of no research suggesting that
holding all studies to a high standard. false belief' or conservation errors occur at certain ages only
One common problem in discussions of the impact of play on when experimenters are unmasked. However, for research on
development is that correlational findings are often discussed as if the benefits of pretend play there are several cases where results
they were causal. When children who play more do better on some obtained with knowledgeable experimenters went away when
other measure, of course it does not mean that the play definitely masked ones were employed (Christie, 1983; Guthrie & Hud-
caused the outcome. Positive correlations between pretend play son, 1979; Pepler & Ross, 1981; Simon & Smith, 1983, 1985;
and a development are only a necessary precondition to pretend P. K. Smith, Simon, & Emberton, 1985; P. K. Smith & Whit-
play being causal. Likewise prominent authors have descdbed ney, 1987). Nonreplications with masked experimenters make a
elaborate pretend worlds they constructed as children, and one strong case for being cautious about pretend play results ob-
might see the earlier behavior as causing their subsequent literary tained with knowledgeable experimenters.
genius (Root-Bernstein, in press), but it is as plausible that their Besides correlational data, nonreplication, and unmasked ex-
creativity led to conjuring up elaborate imaginary worlds at both perimenters, other recurrent problems are very small sample
time points. sizes, nonrandom assignment, confounding implementer with
A second recurdng problem is failure to replicate. For example, intervention (particularly concerning when there is only one
one study shows increases in empathy associated with pretense implementer per condition and interventions last for several
training (Saltz & Johnson, 1974) and another does not (Iannotti, months), control conditions that differ beyond pretend play,
1978), and typically only the positive finding is cited. If other key confounding content with pretend play, and unsound statistical
expedmental factors are essentially equal, either the reported pos- practices like using subsets of data and one-tailed tests without
itive result reflects a Type I error or the failure reflects a Type 11 prior rationale.
error. Inconsistent findings in correlational studies contradict the Methodological problems are so prevalent in this literature that
causal view but would be expected with either the equifinality or meta-analysis is precluded. E. P. Fisher (1992) did a meta-analysis
epiphenomenalism. For equifinality, nonreplications would occur of the impact of play (generally) on development, despite aware-
when the altemate route was stronger in one study, and including ness of these limitations (see "Shortcomings of the Studies," pp.
it masked the effect of pretend play; for the epiphenomenon 164-168), but he also did not have a consistent even-handed
position, nonreplications would occur because the underlying approach to which statistics he included, and further, he used some
cause sometimes accompanied pretend play and sometimes did wrong statistics that inflated his result. As a particularly egregious
not. In the literature extolling play's benefits, failures to replicate illustration of this, from Chdstie (1983) he used the statistic
are often ignored. pertaining to a vadable named vadable {F = 257.67), reflecting
A third problem concerns experimenter bias. Every under- the overall sample scores on five vadables, when the far smaller
graduate research methods course should impart the importance Vadable X Time X Condition statistic (F = 0.49) is what should
PLAY DEVELOPMENT
have been used. Careful reading reveals many more problems, yet any other participant). Another common task is the Torrance
this article is often cited (126 times, Google Scholar, as of May 28, Thinking Creatively in Action and Movement Test or TCAM
2012) as evidence that play helps development (e.g., Bergen, 2009; (Tortance, 1981), which includes several subtests, including hav-
Ginsburg et al., 2007; Wyver & Spence, 1999). ing children move like trees in the wind, and also altemate uses.
Because so many studies in this area are methodologically Correlational studies. Several studies have addressed the
unsound, the current literature base is best suited to a descriptive claim that pretending makes children more creative (Ginsburg et
review, on which we now embark. In each section, we begin with al., 2007; J. L. Singer, 1973) by looking for correlations between
theoretical and construct issues, then review studies. A series of 10 naturalistic play and creativity, since they should exist if more
tables compiles the studies pertinent to each domain or subdomain frequent pretenders have become more creative via their pretend
of development. activity. Of course correlations are not evidence of causation, but
After reviewing the studies, each section concludes by dis- if causation exists, correlations should be consistently found.
cussing the evidence with respect to the three views (summa- Naturalistic classroom play has been categorized differently in
rized in Table 12). In these discussions, we sometimes rely on different studies, but a combination of Smilansky (1968) and
the absence of evidence to support a position. We do this with Parten's (1932) schemes has been used most often (Rubin, 2001).
caution, since one can never prove that a relationship does not Smilansky's scheme (derived from Piaget) divides play into cog-
exist (Altman & Bland, 1995). However, inconsistent correla- nitive categories (functional play when a child repeats motor
tion patterns across studies with similar samples and methods actions on objects, construction play when a child builds things,
and reliable coders are against a causal view. Likewise, when dramatic or symbolic play when the child substitutes an imagined
sound experimental methods yield null effects or even effects world for reality, and games-with-rules like Checkers). Parten's
showing play is less positive than the alternative, this is also scheme is focused on the social aspect of children's play: coop-
relevant. Finally, doubt is also cast on a causal view when erative when children are actively interacting in a common group
masking experimenters or equalizing other aspects of interven- endeavor; associative when they interact but not toward a single
tions nullifies previous findings. common endeavor; parallel when they play sitnilarly but side by
A final note before treading into the evidence concerns the side, with little interaction; solitary independent when they play
"straw person" element of the crucial causal view. When put on alone at their own games; onlooker when they watch others play;
the stand, perhaps few would endorse the position that pretend and unoccupied.
play is crucial (in the sense of essential or vital) for various We found eight studies correlating pretend play and creativity
aspects of development. Yet the quotes with which we opened (see Table 2). Seven concemed preschoolers. Typically children's
and additional references throughout this review show that this play in preschool was coded for 1-5 min per day for a period of 20
stance is taken in the literature, so we consider it here. days or more, and then altemate uses with two to four objects (or
in some cases the TCAM or another test) was administered.
Results were inconsistent.
Nonsodal Cognitive Aptitudes
Johnson (1976) found that, controlling for IQ, amount of social
As was seen in our opening paragraph, many scholars have but not solitary fantasy play was related to fluency. This would
asserted that pretend play produces cognitive benefits. One way suggest that something about the social element, rather than pre-
pretend play could help cognition is by predisposing children to a tending in and of itself, was related to creativity. However, John-
generally playful attitude (Dansky & Silverman, 1973) that could son (1978) later failed to rephcate this finding in a very similar
lead to production of unusual ideas, creative problem solving study, showing no relationship between pretend play (social or
(Vandenberg, 1980), and then to other cognitive aptitudes. This solo) and altemate uses. A different study with the same age range
view is compatible with Freddckson's (2001) broaden-and-build also found no relationship between creativity and social pretend
theory of positive emotions, with play eliciting joy, which in tum play (L. Dunn & Herwig, 1992) but found a negative relationship
leads to a broadening of individuals' thought-action repertoires. between originality of responses and solo pretend play that disap-
Vygotsky's ideas on symbolic and abstract thought, just reviewed, peared when IQ was partialled out. Pellegrini and Gustafson
also suggest how pretend play could assist cognitive abilities. Here (2005) also did not find children's frequency of pretend play to be
we discuss evidence that pretend play assists development in five related to creativity (r = -.18) yet did find frequency of construc-
subdomains: creativity, problem solving, intelligence, conserva- tion play was related. On the other hand, a different study found all
tion, and reasoning. pretend play was significantly related to creativity in a sample of
15 high-IQ preschoolers (Moran, Sawyers, Fu, & Milgram, 1984).
Wyver and Spence (1999) found that particularly fantastical pre-
Creativity tend play was related to semantic creativity (naming all the objects
Although creativity has been operationalized in a number of one could think of) but not figurai creativity (making objects from
ways, in the studies on play it has typically been defined as the shapes); two other categories of pretend play that they coded were
ability to produce original content relevant to a particular task not related to either. The remaining preschool study, Lloyd and
(Wallach & Kogan, 1965). The most commonly used measure of Howe (2003), was not useful regarding pretend play because they
combined it with functional play, which was twice as common as
creativity in this literature is the altemate uses task (R. C. Wilson,
pretend play, but it is worth noting that in contrast to Pellegrini and
Guilford, & Christensen, 1953), in which participants give possible
Gustafson (2005) they found construction play (which was coded
uses for common objects, like a paper towel or a paperclip (Dan-
separately) was not related to creativity. In the eighth and final
sky, 1980a). Task responses are typically coded for fluency (num-
correlational study, looking at somewhat older children, Russ and
ber of uses named) and originality (number of uses not given by
LILLARD ET AL.
Table 2
Studies Examining the Effect of Play on Creativity
Note. Type of study: C = correlational; E = experimental; T = training. Type of play: Play = pretense status unspecified; Const. = construction
play (e.g., blocks); Solo = pretend play alone; SPP = social pretend play; PP = pretend play (social unspecified). +: positive relationship to play;
~ : no correlation ot play = nonplay condition; - = negative telationship to play. Masking: Intetvention (Int) or posttest experimenters (Exp). If
masking status was not specified, we assume experimenters were not masked, since that is the unmarked case. Masking for cotrelational studies is
omitted because it is rarely mentioned, even when it is likely (because play observations occutted several years earlier than testing, for example).
Irrel. = irrelevant.
Grossman-MeKee (1990) examined play with Russ's Affect in for whieh they later named uses (Dansky, 1980b; Dansky &
Play Scale (APS), whieh uses 10 min of play with puppets and Silverman, 1973, 1975; Li, 1978). This fit with Sutton-Smith's
bloeks coding generativity. Alternate uses also taps generativity, (1968) quasi-experimental study in whieh boys eame up with more
so it is not surprising that positive eorrelations were found, both alternate uses for traditional boy toys than did girls (although for
eoneurrently and over time (Russ, Robins, & Christiano, 1999). girl toys, they were equal). To check whether experience with the
The eight eorrelational studies show an inconsistent pattern of objeets, rather than playing with them per se, was important,
relationships that does not support the eausal model. The other Dansky and Silverman (1973) ineluded an imitation group, in
models seek alternate routes or third variables that eould underlie whieh ehildren gained experienee by imitating the experimenter
relationships when they are found. With this particular set of handling the objeets. The imitators gave no more uses than the
studies, no obvious other variable emerges. IQ is sometimes re- controls, suggesting that experienee with the objeets did not ex-
lated to creativity (Johnson, 1976; Wyver & Spenee, 1999) but plain the first results (see also Hughes, 1981, as eited in Hutl,
sometimes not (L. Dunn & Herwig, 1992; Lloyd & Howe, 2003; Tyler, Hutt, & Christopherson, 1989).
Moran et al., 1984; Wyver & Spenee, 1999); Johnson (1978) did Pretend play's effect on creativity could be limited to the objeets
not test IQ. Johnson worked with low-socioeeonomie status ehil- at hand. To test this, Dansky and Silverman (1975) used a different
dren, whereas others involved middle-elass ones, but this eannot set of objeets in the test phase, and here again the play group
explain the diserepant results aeross Johtison's own studies. An- produced more uses, suggesting play's hypothesized effect on
other third variable that fits either view is environment (i.e., types creativity generalizes.
of toys supplied), whieh ean drive the types of play ehildren In the Dansky studies just described, although the theoretieal
engage in (MeLoyd, 1983). Because environment was not mea- rationale coneemed pretend play, it is unelear whether ehildren
sured, we eannot evaluate if this eould underlie the ineonsisteneies. pretended with the objeets or just manipulated them. Li (1978)
Regardless, distinguishing the second and third views requires tested whether pretend play might improve creativity above and
experimental studies. beyond play generally. In the pretend play eondition, the experi-
Experimental studies. Four experimental studies found menter told a short fantasy story, then showed ehildren the stim-
higher associative fluency when ehildren first played with objects ulus objeets and said, "Let's make-believe or imagine that these
PLAY DEVELOPMENT
objects could become anything you want them to be. Play with all on creativity, rather than play having a positive one. Further
of these things" (p. 33). Free play, imitation, and control condi- research with a more neutral control condition is needed.
tions were similar to those used in Dansky's studies. After 10 min, Training studies. Expedmental studies (as defined here)
the same expedmentet administered the altemative uses test with examine short-term change; perhaps play does influence creativity
the three objects used in the intervention and a new fourth object. but requires longer incubation pedods. We located seven longitu-
Significant differences were found for one of three old objects (a dinal play training studies. The first three produced effects that
paperclip), for the make-believe group and the free play group, and wete in the expected ditection, but four othets controlled fot adult
for the new object (a screwddver) only for the make-believe group contact and found play training itself had no effect.
versus the control. Overall, Li's children came up with far fewer Dansky (1980a) compared 36 low-income preschoolers in so-
uses than children had in the previous studies, perhaps reflecting ciodramatic play, free play, and object exploration interventions
population vadance. over 3 weeks, with three 30-min sessions per week. The sociodra-
Taking a different tack, Dansky (1980b) examined whethet matic play training involved enacting pretend play themes like
childten who naturally engage in more pretend play would benefit going on a picnic. Children in the free play group could play as
mote from a play intervention. Children were classified as pte- they wished, and they rarely engaged in pretend play. The explo-
tendets if they engaged in pretense more than 25% of the time, or ration training group explored and discussed objects. All exped-
nonpretendets if they engaged in it less than 5%. They wete then menters were masked. The sociodramatic play group outperformed
assigned to free play or control conditions. Suggesting the class- both other groups on altemate uses.
room classifications had validity, in the free play condition 88% of In a similar study, Feitelson and Ross (1973) compared 24
the ptetendets but only 6% of the nonpretenders pretended with the kindergarteners in play tutoring, music tutodng, toy play without
objects. The altemate uses test was given with a different set of tutoring, and control groups, with ten 30-min interventions over 5
objects, and only the pretenders in the free play condition had weeks. Creativity was measuted with the picture completion sub-
higher fluency. Dansky concluded that play induces creativity only test of Thinking Creatively With Pictures (Tortance, 1966), in
for those who are predisposed to pretend. which childten complete up to 10 pictures in a way that "no one
These expedmental studies suggest that play might have a else will think o f and give each picture a title, and Dog and Bone
causal effect on creativity, at least for children who frequently (Banta, 1970), in which childten make up different routes from a
ptetend (see also Sutton-Smith, 1967). However, in these studies dog to a bone. The play tutoring group's scores increased the most
the expedmenters administedng the creativity test knew which on the number of unusual tities given (but not number of pictures
condition each child was in, and perhaps their expectations influ- ot picture elements) and trended toward better performance on
enced children's responses. With altemate uses, the expedmenter Dog and Bone.
elicits answers until they think a child has run out of possibilities. Wyver and Spence (1999) trained 38 children in three types of
More coaxing might inadvertentiy occur when children ate ex- play based on Patten's social categodes and play type (associative
pected to produce mote uses. In an attempted replication of Dan- fantastical, cooperative fantastical, or cooperative constructive/
sky and Silvetman (1973), P. K. Smith and Whitney (1987) used everyday sociodramatic) for 4 ht over 4 weeks, with pte- and
different expedmenters for the intervention phase and the posttest, posttests of figurai and semantic creativity; they included a no-
with the latter masked to condition. Results showed no significant intervention control. The former two groups increased their fan-
differences—in fact the control condition obtained the highest tastical pretend play, and the third did not increase in any type of
mean score. This finding is consistent with another study using a pretend play. However, only the third group showed increases in
masked expedmenter: The number of uses given was not signifi-
both semantic and figurai creativity. Associative fantastical play
cantly different for children in a play condition (Pepler & Ross,
training improved semantic but not figurai creativity, and cooper-
1981, Study 2).^ A less direct test of possible experimenter effects
ative fantastical play did not improve in either. These results ate
occurs when a different hypothesis is being tested. Two studies
quite mild then for the hypothesis that pretend play increases
tested the hypothesis that focused questioning would elicit mote
creativity, since it did not reliably do so fot the fantasy groups and
uses than would playing with objects (Pellegdni, 1981; Pellegdni
the sociodtamatic group also engaged in construction play. Othet
& Gteene, 1980). The hypothesis was upheld: Focused questioning
expedments in this study tested the tevetse direction by training
led to more uses, and free play was not different from the control.
children in creativity and then observing play; results suggested
Another way to check for expedmenter biasing is to videotape task
delivery, which could encourage standardization; Russ and complex and bidirectional relationships.
Kaugars (2001) did so, and childten who engaged in ptetend play The results of these studies suggest that the play training can
genetated no more uses than children who did puzzles. All this inctease childten's creativity. However, it is unclear whether ex-
suggests that the altemate uses task might be particularly vulner-
able to expedmenter bias.
' Their first experiment, using a knowledgeable experimenter, yielded
A solution is to use a different creativity task, like making a some significant differences, although observing an experimenter play had
collage, the creativity of which is assessed by masked judges as much of an effect on fluency as playing oneself. Their second experi-
ment with a masked experimenter did not replicate the first one but did
(Howard-Jones, Taylor, & Sutton, 2002). Unfortunately the study
yield a near-significant difference in originality (not fluency) of responses
that did this did not use a good control condition. Six- and
(p < .052). Planned comparisons were done, using one-tailed tests, yield-
7-yeat-olds played with salt dough fot 25 min or copied words ing positive results for free play. Pepler and Ross (1981) claimed to have
from the board (and if they finished eatly, were told to statt over). presented evidence that free play promotes divergent thinking, but the
Afterward, they made collages, and the play group's were more inconsistent findings within the study and the use of less rigorous statistical
creative. However, perhaps forced copying had a deleterious effect methods make us view this as an unreliable result.
LILLARD ET AL.
perimenter involvement was greater in the play training group. play is an epiphenomenon of some other relationship to creativity,
Christie (1983) controlled for adult contact (and used a blind coincident with adult interaction.
experimenter, which Dansky did but the others did not) when If pretend play is epiphenomenal, one must explain why corre-
comparing preschoolers in play tutoring with those in skills tutor- lations are found between pretend play, or social pretend play, or
ing, using nine weekly 20-min sessions. With adult interaction construction play, in some studies but not others. Most of the
controlled, both groups improved on the fiuency factor of the correlational studies took place on school playgrounds, hence in
TCAM, with no advantage for play over skills tutoring. There were different and unique settings. Perhaps some feature of the settings,
no changes for either group on the originality or the imagination like varying styles of adult interaction or different toys those adults
factors of the TCAM. Qn the other hand, in this study there was no provide, could explain the inconsistent patterns of correlations.
increase in pretend play in the play training group. One could In closing, note that the studies here were limited in many ways.
argue that this should not be necessary—that pretending within the Ns were often very small, training schedules were often paltry, and
intervention should be all that is needed; the sociodramatic/ the operationalization of creativity was limited. Just one study
construction play children in the Wyver and Spence study im- examined artistic creativity, and although it needs follow-up, it
proved in creativity without increasing their pretend play. Yet they was promising. There is not a compelling case that pretend play
had a 4-hr intervention over 4 weeks, whereas Christie's had 3 hr improves creativity as it has been measured, and there is a strong
over 9 weeks. Perhaps pretend play has an effect but only after a need for more high quality research.
more intensive play intervention schedule.
Yet P. K. Smith and Syddall (1978) and P. K. Smith, Dalgleish, Intelligence
and Herzmark (1981) did increase pretend play with their more
lengthy intervention and also controlled for adult contact, compar- Vygotsky (1967, 1978) is the preeminent proponent of the
ing fantasy play and skills tutoring groups. Although the sample position that pretend play increases intelligence. His claim was that
size was very small in the first study (14), 65 children participated repeated experiences separating object from referent develop ab-
in the second study. They again found no group differences in a stract reasoning. The definition and measurement of intelligencs
creativity outcome, this time on the Dog and Bone task. are controversial (Neisser et al., 1996), but abstract reasoning is
clearly central. In the literature concerning play, intelligence has
One final study involved slightly older children who partici-
been operationalized by a variety of measures, such as Raven's
pated in five 30-min training sessions over 3 to 5 weeks (Moore &
Progressive Matrices, the Stanford-Binet and the Peabody Picture
Russ, 2008). Experimental children were asked to play out partic-
Vocabulary Test (PPVT); which are correlated (Sattler, 1992), and
ular stories, emphasizing affect or fantasy and story coherence. A the Wechsler Preschool and Primary Scale of Intelligence
control group did puzzles, and experimenter contact was standard- (WPPSI). Below we review correlational and training studies.
ized to involve similar levels of encouragement in each group.
Although the play groups did play more following the intervention, Correlational studies. Several studies have examined rela-
tionships between naturally occurring play and intelligence in
the control group actually increased significantly more in fluency
preschoolers, and results have been inconsistent (see Table 3).
on the alternate uses test. Thus although Christie (1983) is incon-
Three studies were previously discussed regarding creativity. In
clusive because the intervention was perhaps insufficient, these
the first of these (Johnson, 1976), social pretend play was related
other three studies suggest that increased adult interaction might
to the PPVT and the Picture Completion subset of the WPPSI, but
drive increases in children's creativity, rather than pretend play
nonsocial pretend play bore no relation. This suggests that being
itself.
able to incorporate others into one's play is related to intelligence,
Summary. The evidence that pretend play enhances creativ- but failing to do so is unrelated.
ity is not convincing. Correlational studies are inconsistent, with The other two studies also discussed for creativity found nega-
some showing relationships only to social pretend play, pretend tive relationships between intelligence and lower levels of play.
play, or construction play, and other studies failing to show rela- Lloyd and Howe (2003) used the same intelligenee tests as John-
tionships to those same constructs. Inconsistent correlations are son (1976) but did not code social play; a negative relationship was
against a causal model and unfriendly but not fatal to equifinality. found with a variable combining Parten's lowest categories (on-
In the experimental studies with strong control conditions, evi- looker and unoccupied). L. Dunn and Herwig (1992) coded social
dence that play increases associative fluency disappeared when play but found no correlation with IQ; instead they found a
experimenters were masked. This shows that the play ethos is a negative correlation to solitary play, such that children who en-
considerable problem in this domain, and it is fatal to both the gaged in more solitary play had lower IQ scores.
causal and equifinality views. Furthermore, effective training stud- An earlier study also found a negative relationship (r = -.19)
ies with controlled adult contact found that skills training increased between IQ and solitary functional play for a sample of 122 older
creativity just as much as pretend play training. Although this 4-year-olds in preschool settings (Rubin, 1982); it also found a
could support equifinality and epiphenomenalism, equifinality was negative relationship for onlooker play and a positive one for
eliminated by the experimental studies with masked experiment- construction play, but no relationship to pretend play. Parten
ers. Unless one argues that the length of the play interventions was (1932), using a much larger age range (1-4 years) and smaller
insufficient (training ranged from 3 to about 26 hr, with no linear sample (n = 42), also found a negative relationship between IQ
relationship to results) or that the creativity tests were invalid, the and solitary play (r = -.20, an equal effect size that was nonsig-
pattern of results from the methodologically sound studies nificant with the sample size) and found larger positive relation-
(masked experimenters, control conditions that equalize important ships with both cooperative and parallel play. Taking these studies
nonfocal factors, effective interventions) best supports that pretend together, a consistent picture that emerges is that in preschool
PLAY DEVELOPMENT
Table 3
Studies Examining the Effect of Pretend Piay on Inteiiigence
Masked Masked
Type Citation + - Int Exp Notes
C Johnson (1976) SPP Did not control for age, wide span
C Lloyd & Howe (2003) Solo No play Social not coded
C L. Dunn & Herwig (1992) SPP Solo
C Rubin (1982) Const. F
C Parten (1932) SPP Solo
C Peisach & Hardeman (1985) PP
C Cole and Lavoie (1985) SPP, PP
C Johnson et al. (1982) Const. Solo, SPP F
T Saltz et al. (1977) SPP, D No No Did not control for pretest
T Christie (1983) PP = Skills No Yes Adult contact controlled
T Saltz & Johnson (1974) PP = Classify No No Dimension training slightly more effective
than PP training alone
T Schellenberg (2004) D Yes Music > Drama = No lessons
T P. K. Smith & Syddall (1978) PP = Skills Yes No Adult contact controlled
T P. K. Smith et al. (1981) PP = Skills No Yes Found for one test, at one site
Note. Type of study: C = correlational; T = training. Type of play: Const. = construction play (e.g., blocks); Solo = pretend play alone; SPP = social
pretend play; F = functional play; PP = pretend play (social unspecified); D = enacting stories with dolls or other children. +: positive relationship to
play; —: no correlation or play = nonplay condition; - = negative relationship to play. Masking: Intervention (Int) or posttest experimenters (Exp). If
masking status was not specified, we assume experimenters were not masked, since that is the unmarked case. Masking for correlational studies is omitted
because it is rarely mentioned, even when it is likely (because play observations occurred several years earlier than testing, for example).
settings, where other children are available to play with, choosing observed support this possibility, level of intelligence might
to play alone or not play at all might be indicative of lower induce level of play, rather than the reverse. Training studies
intelligence. can resolve this.
Three other studies could be seen as going against this Training studies. We found six training studies that exam-
picture, but methodical variations can explain the difference. ined intelligence. Just one reported a solid and unique pretend play
Peisach and Hardeman (1985) scored fantasy play either via training effect (Saltz, Dixon, & Johnson, 1977). Three cohorts of
observation (4-year-olds) or interview (5- to 7-year-olds) and children were given, over 6 months, a 15-min intervention 3 times
administered a test similar to the Raven's; they did not obtain per week enacting thematic fantasies (like The Three Biiiy Goats
significant positive correlations. However, the small number of Gruff), discussing those same stodes, or acting out typical expe-
4-year-olds (10) and the fact that older children were inter- dences like going to the doctor. A control group engaged in other
viewed rather than observed in school settings could be why. preschool activities. The sociodramatic and thematic fantasy
Cole and Lavoie (1985) found no relationship between solitary groups had higher posttraining intelligence scores than the other
or social pretend play and PPVT scores in children ages 2 to 6, groups. However, the researchers who did the training also adtnin-
but in this study dyads were randomly paired, which might have istered the posttests, which could be problematic for studies of
affected the nature of their play. Johnson, Ershler, and Lawton pretend play. Among the other five studies, three of which used
(1982) found that 4-year-olds who engaged in more functional masked posttesters, pretend play tutodng showed no more benefit
play at preschool scored lower on the PPVT and the Raven's, than skills/classification tutodng (Chdstie, 1983; Saltz & Johnson,
and those who engaged in more construction play scored higher. 1974; P. K. Smith et al., 1981; P. K. Smith & Syddall, 1978) and
Although this is consistent with the picture painted thus far, its less benefit than music (especially voice) training (Schellenberg,
third finding is not: Contra Johnson (1976), sociodramatic play 2004); the only hint of a difference favoring play was in one study
frequency was not related to intelligence (cf. Henniger, 1991). (P. K. Smith et al., 1981), on one of eight tests (geometdc design),
Perhaps the preschool setting was influential here: It was a in one of two schools, which is not better than would be expected
cognitive-based program, and almost three times as much con- by chance. Thus when more sound methods were used, pretend
struction as pretend play was coded. Whereas in some settings play training did not increase intelligence more than the compar-
pretend play is most frequent (Parten, 1933), here it was the ison condition. Regarding Schellenberg (2004), however, we have
least common type of play. It might be the case that the program two concems: differential dropout (17% in the keyboard condition)
offered abundant construction materials and few toys inspiring and combining the keyboard and voice groups for analysis when
sociodramatic play, restricting its range and preventing stronger the IQ gain of the drama group was only one point less than the
correlation. gain of the keyboard group.
The picture that emerges is that less intelligent children Summary. Relationships are found in natural settings be-
engage in lower levels of play in most of the preschool settings tween levels of play and intelligence, but the direction of effects is
studied. By the Vygotskian model, these children failed to uncertain. Training studies with solid methods suggest that pretend
become more intelligent because they did not engage in higher play is no better than other adult interventions in raising intelli-
levels of pretend play. Although the consistent correlations gence scores, and music interventions raised them more (cf.
10 LILLARD ET AL.
Wenner, 2009). Perhaps some feature of adult interaction or some examine causality and to more specifically address the issue of
aspect of music training drives group differences, which is more pretend play.
consistent with epiphenomenalism. Ideally in future research one Experimental studies. In the original study to use the lure-
could remove these key features, or isolate them for singular retrieval paradigm with children, Sylva (1974, summary published
presentation in a third condition. in Sylva, Bruner, & Genova, 1976) gave preschoolers a 1-min
demonstration of how to put a clamp on a single stick. Then one
Problem Solving group was given 10 sticks and seven clamps and allowed to play
for 9 more min, an observe group was given an additional 1-min
Problem solving involves inventing strategies to overcome an
demonstration in which the adult joined two sticks with a clamp,
obstacle and reach a goal. Theorists have speculated that playing
and a control group went straight to the lure-retrieval problem. The
with objects enables children to think about them in myriad ways
problem involved removing a chalk from a plexiglass box by
and thus employ them in new ways to solve problems. Studies of
clamping two sticks together, using them to dislodge the chaUc, and
problem solving and play have typically used Köhler's lure-
then raking it toward oneself. When children did not engage with
retrieval paradigm, in which one must join two sticks together with
the tools during the problem time they received up to five hints,
a clamp or block in order to reach something. With this paradigm,
culminating in direct instruction.
it is not always clear if children are pretending: mapping an
altemate reality on to the objects. But in discussions of the studies, The play and observe groups both outperformed the controls,
pretend play is often considered key (Garvey, 1990). See Table 4. with about 40% of each obtaining the chalk with no hints, and
Correlational studies. Four correlational studies of play and there were also no differences between these groups in latency to
problem solving found that engaging in construction but not pre- solve the problem. Although overall children in the play group
tend play in preschool was positively related to performance on the solved the problem with fewer hints by a McNemar's test, exam-
lure-retrieval problem (Cheyne & Rubin, 1983; Pellegrini & ination of Table 6 in Sylva's thesis shows the play group received
Gustafson, 2005; Rubin, 1982), although for one study that rela- 74 hints, the observe group 77, and the control group 120, so the
tionship held only for boys (Gredlein & Bjorklund, 2005). Con- difference between the first two groups was utterly trivial. We
struction play could be related to performance on the lure-retrieval conclude that playing with a set of objects for 9 min, having had
problem due to a third variable. For example, if some children are a 1-tnin demonstration of how an object works, is as good as a
particularly drawn to constructing things, this could lead to high 2-min demonstration of how that object works. Another study
scores on both measures. Experimental studies are needed to removed demonstration altogether and found the effects of 10 min
Table 4
Studies Examining the Effect of Play on Problem Solving
Masked Masked
Type Citation Int Exp Notes
Note. Type of study: C = correlational; E = experimental. Type of play: Play = pretense status unspecified; Const. = construction play (e.g., blocks);
PP = pretend play (social unspecified); Obs = observe. +: positive relationship to play; ~ : no correlation or play = nonplay condition; — = negative
relationship to play. > = did better than. Masking: Intervention (Int) or posttest experimenters (Exp). If masking status was not specified, we assume
experimenters were not masked, since that is the unmarked case. Masking for correlational studies is omitted because it is rarely mentioned, even when
it is likely (because play observations occurred several years earlier than testing, for example). Irrel. = irrelevant.
PLAY DEVELOPMENT 11
of play versus a control were task dependent, appearing for a aeross play and observe eonditions. This pattern of results eould
lure-retrieval but not a dislodging problem (Vandenberg, 1981). support epiphenomenalism or equifinality as regards eonstruetion
A further study using the lure-retrieval paradigm addressed the play. But the faet that eonstruetion and not pretend play is the
differential exposure time aeross Sylva et al.'s conditions by consistent eorrelate to solving the lure-retrieval problem, and that
extending the observe eondition so it would also take 10 min (P. K. pretend play even interferes with using objeets as tools, suggests
Smith & Dutton, 1979). To further test problem solving, they that pretend play does not help problem solving, at least not the
added a seeond test, in whieh the ehild had to put three stieks types used in research thus far. Researeh does show that exploring
together to reach the reward. Performanee aeross the observe and speeially eonstrueted puzzle toys leads to figuring out their solu-
play groups was equal on the first problem by all three measures— tions, and further researeh should examine if sueh "play" helps
percentage solving, solution latency, and number of hints. How- problem solving generally.
ever, the play group solved the seeond problem more quickly than
the observe group, thus they obtained a play-favored result for one Reasoning
of six measures. Yet Smith and his eoUeagues were leery that
experimenters might have introduced bias into the procedure, so Another cognitive skill sometimes diseussed with referenee to
they reran the study with a masked experimenter, and eondition the purported benefits of pretend play (D. G. Singer & Singer,
differenees were eliminated (Simon & Smith, 1983). They con- 1992, p. 237) is solving logical syllogisms, in whieh one must
eluded that experimenter bias in delivering hints in the prior study reason from false premises like, "Dogs live in trees. Rex is a dog.
led to its positive results, a hypothesis eonfirmed by a later study Does Rex live in a tree?" The logically eorreet answer is that he
(P. K. Smith et al., 1985; see also Simon & Smith, 1985). Null does, but the problem is difficult sinee one must set aside one's
results for play in a further study led to the conclusion that there real knowledge. Several studies (see Table 5) have found ehildren
should be "serious questions about the merits of play for enhaneing do better on sueh problems when they are embedded in fantasy
problem solving" (Vandenberg, 1990, p. 271). (Dias & Harris, 1988, 1990; Hawkins, Pea, Gliek, & Seribner,
1984; Kuezaj, 1981; Riehards & Sanderson, 1999), for example by
Finally, there is also controversy regarding the fantasy versus
using exaggerated pretend intonation, or saying, "Let's pretend
construction play element in all of these studies. Sylva (1977)
that everything in the stories is true" (Dias & Harris, 1988, p. 210).
reexamined her results and found ehildren who pretended with the
However, Harris and Leevers (2000) thought this might be beeause
stieks were most apt to solve the problem, yet in an explicit study
fantasy manipulations got ehildren to eonsider the premises more
of this (Hughes, 1981, 1983, as described in Hutt et al., 1989),
carefully. A series of experiments showed that using any eue that
ehildren who pretended with the stieks needed more hints to reach
"elarified the experimenter's intention that the ehildren should
a solution than ehildren who had engaged in exploratory/
aeeept the premises as a basis for reasoning" (p. 77) helped. This
constructive play (see DeLoache, 2000, for another ease in whieh
supports the equifinality view; Pretend play, as operationalized in
play impedes cognition). From this focused study, construction but
these studies, is one of many means to enhance children's ability
not pretend play appears to be the important factor.
to solve logical syllogisms.* It makes sense that pretend play might
Some reeent experiments have shown that being allowed to help ehildren reason about false premises, sinee they are defini-
freely interact ("play" in the sense of explore, an action with whieh tional to pretend play; One aets as if something false were true.
play is often contrasted; see Berlyne, 1960; Hutt et al., 1989, p. 11) Further researeh should use correlational and training paradigms to
with specially designed puzzle toys known to do interesting things explore whether pretending affeets logical reasoning more gener-
ean lead to discovering solutions that are more effieient than a ally. We note that this reasoning researeh is also consistent with
single inefficient solution that was taught (Bonawitz et al., 2011; epiphenomenalism; researeh separating pretend play from the eue
Buehsbaum, Gopnik, Griffiths, & Shafto, 2011). In these eases, the to eonsider the premises is needed to show whether pretend play
speeifie problem—getting a specially eonstructed toy to make alone is effeetive.
musie—is probably important; there is no indication that explora-
tion improved problem solving generally, although sueh a study
would be welcome. Reviews and meta-analyses of the "diseovery Conservation
learning" literature have eoneluded that ehildren tend not to learn
what was intended in open diseovery learning paradigms (Alfieri, Conservation, or understanding that objeets retain eertain eore
Brooks, Aldrich, & Tenenbaum, 2011; Mayer, 2004). Rather, properties after superfieial transformations, is structurally like
ehildren tend to learn best when problems are structured sueh that pretending; In both eases children must keep one reality in mind
ehildren are likely to find the solution (Klahr & Nigam, 2004) as while simultaneously focusing on a represented alternative (Fink,
they are in Montessori edueation. This is in line with a "playful 1976). One must think about the water in a short glass while
learning" (Hirsh-Pasek et al., 2009) but not an open free play looking at that same water in a tall glass, or about the telephone
approach (see Vandenberg, 1990). while holding the banana. However, studies do not support a
natural relationship between pretend play and eonservation (see
Summary. Correlational studies show that construction but
Table 6), since three correlational studies failed to detect one
not pretend play is correlated with solving problems that involve
(Aisenson, 1978; Doyle, Cesehin, Tessier, & Doehring, 1991;
eonstruetion. Does the construction play eause the problem solv-
ing, or does a eommon trait like a propensity toward constructing
underlie both (see Pellegrini, 2009)? Experimental studies eould * Similarly, for philosophers. Twin Earth (Putnam, 1975/1996), an imag-
shed light on this, but they have had ineonsistent and mild results inary place that is just like real Earth but removed from it, has provided
that have not replieated with more tightly controlled procedures or fertile grounds for teasoning from hypothetical premises. Lillard (2001a)
masked experimenters; in those eases performanee has been equal discussed at length how pretend play is like a child's Twin Earth.
12 LILLARD ET AL.
Table 5
Studies Examining the Effect of Pretend Play on Reasoning
Masked Masked
Type Citation Int Exp Notes
Note. Type of study: E = experimental. Type of play: PP = pretend play (social unspecified), -t- : positive relationship to play; ~ : no correlation or play =
nonplay condition; - = negative relationship to play. Masking: Intervention (Int) or posttest experimenters (Exp). If masking status was not specified, we
assume experimenters were not masked, since that is the unmarked case. Masking for correlational studies is omitted because it is rarely mentioned, even
when it is likely (because play observations occurred several years earlier than testing, for example).
Johnson et al., 1982). These are not discussed, but they ate prob- could have both a teal and a pretend identity. The play children
lematic for a causal account. Training studies have had mixed subsequentiy did better on standard conservation tasks. Two meth-
results, with positive ones eventually explained by othet factots. odological problems hold for this and two further studies from
The fitst ttaining study had partially positive results. Fink Golomb's laboratory (Golomb & Bonen, 1981; Golomb, Gowing,
(1976) encouraged one group of 12 to pretend, gave another group & Fdedman, 1982). First, the trainers were the posttestets; Guthde
extra free play time but did not interact, and also had a no- and Hudson (1979) failed to replicate the result with masked
intervention control group. Interventions lasted 25 tnin, twice a expedmenters. Second, better performance might have been due to
week for 4 weeks; masked expedmentets administered the post- the extensive and pointed questioning about how play objects
tests. The pretend play group showed a trend toward improvement could simultaneously have two identities. Golomb et al. (1982) in
in quantity and number conservation and significant improvement fact cited an unpublished study (Golomb & Adams, 1978) as
in social role conservation (understanding that a father was also a showing that verbal inquity was key (see also Hughes, 1981, as
doctor). However, adult involvement was very different across the cited in Hutt et al., 1989). Further, these later Golomb studies
conditions and might be the source of the effects. showed that conservation ttaining was as ot mote effective than
In the next ttaining study on play and conservation, Golomb and ptetend play ttaining.
Comelius (1977) assigned 4-yeat-olds to symbolic play training or In sum, there is no compelling evidence that ptetend play leads
a construction play control condition. The play training involved to conservation. The crucial causal view requires consistent cor-
engaging children in pretense and then asking how the play objects relations that do not exist. Equifinality suffers from this lack of
Table 6
Studies Examining the Effect of Pretend Play on Conservation
Masked Masked
Type Citation -1- ~ Int Exp Notes
Note. Type of study: C = correlational; T = training. Type of play: SPP = social pretend play; PP = pretend play (social unspecified); ConT. =
conservation training. -H: positive relationship to play; ~ : no correlation or play = nonplay condition; — = negative relationship to play. Masking:
Intervention (Int) or posttest experimenters (Exp). If masking status was not specified, we assume experimenters were not masked, since that is the
unmarked case. Masking for correlational studies is omitted because it is rarely mentioned, even when it is likely (because play observations occurred
several years earlier than testing, for example).
PLAY DEVELOPMENT 13
Table 7
Studies Examining the Effect of Piay on Theory of Mind
Masked Masked
Type Citation Int Exp Notes
Note. Type of study: C = correlational; CL = cross-lag correlation; E = experimental; T = training. Type of play: Const. = construction play (e.g.,
blocks); Solo = pretend play alone; SPP = social pretend play (including imaginary companion play in Taylor studies); PP = pretend play (social
unspecified); D = enacting stories with dolls or other children. +: positive relationship to play; ~ : no correlation or play = nonplay condition; - =
negative relationship to play. Masking: Intervention (Int) or posttest experimenters (Exp). If masking status was not specified, we assume experimenters
were not masked, since that is the unmarked case. Masking for correlational studies is omitted because it is rarely mentioned, even when it is likely (because
play observations occurred several years earlier than testing, for example). AR = appearance-reality; PT = perspective taking; IC = imaginary companion;
ToM = theory of mind.
(Rubin & Maioni, 1975),'' but two did not (Cole & LaVoie, 1985; sistent results for it. Studies relating social cognition tasks of this
Peisach & Hardeman, 1985); yet the latter found a positive rela- earlier era to those of today would be useful.
tionship to moral perspective taking. For affective perspective More recent studies have operationalized social pretend play in
taking, one had a positive relationship to social pretend play two ways, broadly speaking: (a) impersonation, including role
complexity but not amount (Connolly & Doyle, 1984), but two enactment and having an imaginary companion, and (b) joint
found no relationship (Cole & LaVoie, 1985; Rubin & Maioni, proposals and role assignments, in which a child states what they
1975). Connolly and Doyle (1984) found no relationship to cog- are pretending. We located 12 studies relating these behaviors to
nitive perspective taking (matching gifts to recipients, from Fla- ToM (Astington & Jenkins, 1995; Doyle & Connolly, 1989; J.
vell, Botkin, Fry, Wdght, & Jarvis, 1968). Thus earlier studies of Dunn & Cutting, 1999; Goldstein & Winner, 2010; Hughes &
this issue were inconsistent, with three positive relationships and
six null ones. One possible reason for some inconsistency is that
Borke's affective perspective-taking task could rely on vocabulary 'This article itself is ambiguous as to whether the "dramatic play"
more than deep understanding, reflecting a language-pretend play category was social, but elsewhere Rubin discussed this study as revealing
relationship. Spatial perspective taking is influenced by familiadty relationships to "symbolic or dramatic play with peers" (Rubin & Pepler,
with task matedals (Borke, 1975), and perhaps this led to incon- 1980, p. 224).
PLAY DEVELOPMENT 15
Dunn, 1997; Lalonde & Chandler, 1995; Lindsey & Colwell, 2003; creased their pretending. As an afterthought she also looked at
Nielsen & Dissanayake, 2000; Schwebel et al., 1999; Taylor & their verbal skills. She did not test for any of the social cognitive
Carlson, 1997; Taylor, Carlson, Maring, Gerow, & Charley, 2004; outcomes that she theorized pretense might infiuence (cf. Bums &
Youngblade & Dunn, 1995), and findings were inconsistent. For Brainerd, 1979), but many others have done so, making social
example, some studies found relationships only with specific tasks, cognition the most studied potential outcome of pretend play
like with appearance reality but not false belief (Schwebel et al., training. Yet all but one study had a serious methodological
1999), or with role enactment but not explicit role assignment shortcoming preventing clear conclusions: The experimenters
(Youngblade & Dunn, 1995), or with joint proposals and role were not masked, or the level of adult contact across conditions
assignments but not social pretend play generally (Astington & was very different, or the training explicitly "taught to the test."
Jenkins, 1995). Several studies used a large age range but did not Some studies have additional problems as well. The most solid
partial out age, when both social pretend play and ToM scores study had null results.
increase with age (Taylor & Carlson, 1997). With age partialled The first six studies we discuss lacked masked experimenters,
out, in some studies the association became a trend (Doyle & which we now know is problematic in studies of the effects of
Connolly, 1989) or disappeared entirely (Hughes & Dunn, 1997). play. Saltz and Johnson (1974) found improvement in affective
Other studies seem less solid because matemal education was very perspective taking following thematic fantasy training (acting out
strongly related to both social pretense and ToM, and socialization stories) but not following training in identifying object dimensions
practices could undergird both outcomes independently (J. Dunn & or no training at all. However, Saltz's second study (Saltz et al.,
Cutting, 1999). 1977) showed no main effect of thematic fantasy or sociodramatic
However, Youngblade and Dunn (1995) used a limited age play, yet they went on to compare thematic fantasy only with the
range and partialled out verbal ability, and Astington and Jenkins other three groups combined (sociodramatic play, fantasy discus-
(1995) partialled out both age and verbal ability; both studies still sion, and control), and only for the first two years of the 3-year
found relationships between aspects of social pretense and ToM; study—not accepted statistical practice.
Schwebel et al. (1999) partialled both factors and still found a Bums and Brainerd (1979) engaged children in sociodramatic
relationship but only to appearance-reality. Although it is incon- play or construction play, preceded and followed by a variety of
sistent, the relationship appears often enough that it is frequently perspective taking tasks. A summed perspective taking score re-
remarked upon in the literature (Harris, 2000; Kavanaugh, 2006; vealed a significant effect of sociodramatic relative to construction
Lillard, 2001a). It is important to consider direction of effects. play, and each improved more than the no treatment group. How-
Longitudinal studies, of which we found two, can shed light ever, effects were inconsistent across tasks. A doctoral student in
here. Youngblade and Dunn (1995) assessed pretending at 33 Brainerd's laboratory went on to do a more rigorous and larger
months and ToM at 40 months and found that role enactment (but replication and extension of this study (Scheffman, 1981, de-
not total amount of social pretending, diversity of pretend themes, scribed in Brainerd, 1982). As summarized by Brainerd, "when it
or explicit verbal role assignment) predicted later ToM. The role comes to the question of how powerful dramatic (and constructive)
enactment finding is consistent with simulation theory and a play is as an enhancer of conceptual knowledge, the results were
pretend-to-ToM causal view. In a study cited far less often (Google extremely disappointing. For all intents and purposes, there were
Scholar, March 19, 2012), Jenkins and Astington (2000) tested no learning effects" (p. 125).
children ages 34-45 months three times—at an initial point, then Two more recent studies (Goldstein & Winner, 2010, 2012)
3.5 and 7 months later—for ToM, language, and the proportion of examined whether acting classes (compared to other arts classes)
their social pretend play turns dedicated to making joint proposals improve ToM and empathy; they lacked random assignment as
in pretense and explicitly assigning roles. Here, earlier ToM pre- well. Children who want to take acting classes might differ from
dicted later social pretend play, but not the reverse. The different other children in ways that infiuence ToM development, and there
directions of effects seen in these two studies can be explained in were in fact group differences in some key measures prior to
at least four ways: (a) The first study did not allow for finding the training. In addition, although positive results were found on some
ToM-to-pretend direction, since it did not include ToM tests at tests, results were inconcordant across the experiments.
earlier time points; (b) the first study used verbal and behavioral A sixth unmasked training study also had unequal adult contact
indicators of pretending, whereas the second used only verbal across conditions (Dockett, 1998). The experimental group visited
ones; (c) only role enactment is related, and the second study did a pizza restaurant and then a pizza play area was set up in their
not measure it; (d) the first study used younger children. More classroom. Over the next month, an experimenter encouraged their
longitudinal studies examining a range of variables are needed to play toward more complex levels. Teachers added props and
more definitively determine the direction of effects between social resources to promote pizza play, posted a photographic record of
pretend play and ToM. Yet still a third variable could underlie the play on the walls, and devoted large blocks of time to it. From
results found with correlational methods; training studies provide pre- to post- and even on a delayed posttest, the play group
the best test. improved more on ToM.
Training studies. Many pretend play-social cognition train- Some 1970s-era training studies also had unequal adult contact
ing studies were conducted in the 1970s (see summary in Rubin, across experimental and control groups but did use masked exper-
1980), inspired by Smilansky (1968). Noting that middle-class imenters. Fink (1976) found that sociodramatic play led to in-
children play more than lower class ones and show many of the creases in social role but not spatial perspective taking, and Ros-
advantages pretense might be expected to confer, Smilansky en's (1974) sociodramatic play group improved significantly more
trained children of lower class immigrants to Israel to engage in on perceptual and semantic role taking. In this latter study the
sociodramatic play for 90 min/day. Nine weeks of training in- trained group had four times more adult contact.
16 LILLARD ET AL.
P. K. Smith and Syddall (1978) explicitly equalized experi- erally (Bretherton, 1989; Fein, 1989; Fein & Kinney, 1994). The
menter contact for two tutoring groups—skills (jigsaw puzzles, seeond theory highlights soeial pretend play because it requires
games) and soeiodramatic play—although experimenters were not negotiation (Harris, 2000; Lillard, Pinkham, & Smith, 2011). Stud-
masked. Cognitive perspective taking improved significantly more ies eoneemed with social skills tend to measure children's behav-
in the play group. Smith and Syddall then questioned the tasks they ior in real-life settings with peer soeiometrie ratings and teaeher or
(and Rosen, 1974) had used. One was the gift assignment task and parent reports.
the other was putting objects with the appropriate person, for
example a stethoseope with a doetor. Sueh tasks require simple Solitary Pretense
matching paralleled in the soeiodramatie play training, when props
are assigned to roles. Three other early studies examining the Addressing the first theory are five studies correlating soeial
influenee of role play training on soeial eognition, although quite skills and solitary pretense (see Table 8). The first two were
sound in most ways, also seem to "teaeh to the test" (Chandler, playground studies using peer soeiometrie ratings; one found no
1973; Chandler, Greenspan, & Barenboim, 1974; Iannotti, 1978). relationship (Rubin, 1982), and the other found a negative one
Beeause of his eoneem about this issue. Smith conducted a (Rubin & Daniels-Beimess, 1983). However, as has been noted, on
follow-up study with a more diverse array of soeial eognitive tasks a school playground the norm is soeial pretend play, and ehildren
and this time also employed a masked posttester and equalized who instead play alone might already have poor soeial skills. Thus,
contact (P. K. Smith, et al., 1981). Using these very sound meth- these results do not really speak to whether solitary pretend play
ods, they found no improvement in role taking in the play tutoring might develop soeial skills.
group or the skills training group. The authors concluded that some The other three studies provide a better test. Two found a
aspect of adult contact, rather than pretense, was responsible for positive eorrelation between children's solitary pretend play in a
prior findings. One might fault this study in that one of two settings laboratory setting and a teaeher measure of soeial eompetenee
showed no posttraining inerease in fantasy play, but it still ought (MeAloney & Stagnitti, 2009; Uren & Stagnitti, 2009), but the
to have obtained results at the setting where fantasy play did third (Swindells & Stagnitti, 2006) found no relationship using a
inerease. Thus the one study in the group without a significant parent measure. Close examination of the measures used in the
methodologieal problem showed no significant inereases on ToM first two experiments reveals that soeial compliance, tapped by
measures for either pretend or skills tutoring. both the teaeher report and pretend play measures, might undergird
the relationships. Further researeh should examine this possibility.
Summary
Social Pretense
Logically, it seems like pretend play eould well assist ToM, but
solid evidence that it does so is lacking. Correlational results have If soeial pretend play helps develop soeial skills, one would
been ineonsistent, and one study has suggested a reverse direction expect that (a) one would see more socially competent behavior
of effeets: A more developed ToM enables soeiodramatie play. within soeial pretend play than outside of it, (b) more advaneed
Many training studies have been eondueted, but most had at least social pretend play, whether occurring naturally or after training,
one serious methodologieal shortcoming, and there are also im- would predict more advaneed soeial skills, and (e) training in
portant failures to replieate. The most solid study showed no social pretend play would improve social skills.
improvement in ToM from either skills or pretend play training. Does soeial pretend play entail more advaneed soetal behav-
The ineonsistent correlational results and hints of a reverse direc- ior? We found five studies examining this. Two found that
tion of effeets lead us to see the body of evidenee as being more social pretend play involved greater soeial skills than soeial non-
supportive of epiphenomenalism than equifinality. Children who pretend play. In a study eomparing soeial pretend and literal
have more advaneed ToM skills often also engage in more ad- activities, pretense interactions lasted longer, involved larger
vanced pretend play, but possibly beeause of some third variable, groups of ehildren, and had more positive and less negative affect
like having parents (or experimenters) who interact with them in (Connolly & Doyle, 1984). The second study replieated this and
ways that eneourage mental state reasoning and pretend play. also found more reeiproeity/complexity in soeial pretend play
(Connolly, Doyle, & Reznik, 1988).
Social Skills The third study occurred in the laboratory and only involved girl
dyads; it had mixed results. Soeial pretend play involved more
Next we examine evidenee that pretend play improves chil- verbal exchanges and shared foeus than soeial literal play, but it
dren's soeial skills. Soeial skills are distinguished from ToM in also involved longer, less often resolved eonfliets (de Lorimier,
that they involve enactment but not necessarily knowledge. Al- Doyle, & Tessier, 1995). The fact that girl dyads were more able
though soeial competence is empirically assoeiated with ToM to solve their eonfliets outside of the pretense setting might suggest
(Bosaeki & Astington, 1999; Watson, Nixon, Wilson, & Capage, that they were more soeially skilled in literal settings.
1999), knowledge ean be inert; alternatively, one eould eonceiv- Against the idea that soeial pretend play entails more soeial
ably act in a soeially skilled manner without underlying ToM skills, Howes and Matheson (1992) found preschoolers' level of
knowledge. Two theories guide researeh in this area. Both suggest soeial eompetenee was equal aeross soeial and nonsoeial pretense
that pretend play eauses soeial skills; one claims that even solitary contexts, and Doyle, Doehring, Tessier, de Lorimier, and Shapiro
pretending does so (Stagnitti & Unsworth, 2000) beeause ehildren (1992) found the eomplexity and positive affect (whieh could
often pretend about emotional and difficult issues; working index prosoeial interaetion) of peer exchanges to be equal in
through such issues in pretense would enhance soeial skills gen- pretense and nonpretense eontexts, although the latency to com-
PLAY DEVELOPMENT 17
Table 8
Studies Examining the Effect of Play on Social Skills
Masked Masked
Type Citation Int Exp Notes
Note. Type of study: C = correlational; O = observational studies comparing means rather than correlating; E = experimental; T = training. Type of
play: Solo = pretend play alone; SPP = social pretend play; PP = pretend play (social unspecified). -I-: positive relationship to play; ~ : no correlation
or play = nonplay condition; - = negative relationship to play. Masking: Intervention (Int) or posttest experimenters (Exp). If masking status was not
specified, we assume experimenters were not masked, since that is the unmarked case. Masking for correlational studies is omitted because it is rarely
mentioned, even when it is likely (because play observations occurred several years earlier than testing, for example).
plex interaction was shorter in ptetense (Doyle et al., 1992). A betg, 2004). The purpose of the study (mentioned earlier) was
sttength of this latter study is that it conttoUed for wbether children actually to see if music lessons enhance IQ, and parent assessments
were frequent or infrequent pretenders. Dyads in which both of children's social competence (adaptability, social skills, and
childten were classified in another setting as frequent pretenders leadership) were administered fot explotatoty teasons. Six-yeat-
engaged in mote complex social interaction than dyads in which olds wete randomly assigned to 36 weeks of keyboard lessons,
both children were classified as infrequent pretenders, which could voice lessons, drama lessons, or no lessons. Consistent with the
explain the results of studies that found more complex interactions hypothesis, children who received keyboard and voice lessons
in ptetend play: It could be that more socially competent childten increased most in IQ. However, the drama group unexpectedly
ate mote apt to engage peets in social ptetend play, in which case increased most in social competence. It is not clear whether this
the ditection of effects is opposite to that speculated. This leads to result stems from playing out toles ot from the social interaction
the next issue. inherent in doing so vetsus doing keyboard or voice lessons. Still
Does more advanced social pretend play predict greater the findings are very interesting, and future research should clearly
social skills? The findings on this ate also tnixed, with two explore them with control conditions matched for social contact.
studies finding it to be the case, two finding it not to be, and two
tevealing diffetent results by gendet. Rubin and Maioni (1975) Summary
found a positive correlation between sociometdc ratings and social
pretend play in 3- and 4-year-olds. Connolly and Doyle (1984) Results from studies examining correlations between social
replicated this and also found a positive relationship with teachet skills and pretend play are inconsistent, which is against the crucial
reports of social competence with peets, but not with the social causal position. The serendipitous finding in a ttaining study
skill of following teachers' tules. In conttast, Rubin and Daniels- suggests a potential causal relationship from drama training to
Beimess (1983) did not find a positive relationship between so- social skills, which would (along with inconsistent correlations) be
ciometdc status and social pretend play, and Galyet and Evans in keeping with equifinality, but altemate epiphenomenal expla-
(2001) failed to find a positive relationship to teachet-tepotted nations must be ruled out (e.g., perhaps level of interaction was
social skills. Finally, Lindsey and Colwell (2003) and Colwell and diffetent in the dtama ttaining). Pretending is pethaps a route to
Lindsey (2005) had gender-specific findings, with mote mixed-sex social skills, but without mote convincing evidence it is equally
play predicting higher same-sex sociometdc and teacher-reported feasible that social pretending and social skills emerge from a
social competence ratings for gitls but lower ones for boys. This latent factor like sociability or some aspect of interaction.
result is also against the view that pretense leads to bettet social
skills, since boys who ptetend witb gitls should gain social skills Symbolic Understanding: Early Language
just as well as boys who pretend with boys.
Does social pretend play training improve social skills? In ptetend play childten use one object to stand for another,
We located only one study that trained children in social ptetend hence, a symbol. Language is also symbolic (Piaget, 1962; Wemer
play (dtama lessons) and then examined social skills (Schellen- & Kaplan, 1963). Aligning with the causal position, some theodsts
18 LILLARD ET AL.
have claimed that repeated practice using symbols in pretend play number of categories of speech represented (Tamis-LeMonda &
"contributes greatly to language development" (Miller & Almon, Bomstein, 1994). Two studies also found the reverse direction of
2009, p. 63; see also Ervin-Tripp, 1991). Yet perhaps pretend play effects, fi-om language to play (Tamis-LeMonda & Bomstein,
is just one of many behaviors that can improve children's language 1994; Ungerer & Sigman, 1984), although for the former the
or is epiphenomenal to some other factors that give rise to lan- finding disappeared when mother behaviors were partialled out.
guage. In this section we examine evidence for these views with Note that there is some methodological bias against finding rela-
regard to early language, namely, acquiring first words and syntax. tionships from language to play: One study did not test play at the
We end with a brief note about the research on how pretend play older ages (Lyytinen et al., 1999), and Ungerer and Sigman used
affects written language or literacy. Findings are summarized in just one language measure at the earlier age versus five later
Table 9. (although they did find a positive relationship for that one mea-
sure). Finally, a fifth longitudinal study tested slightly older chil-
dren—20 and 28 months—and did not find that play predicted
Correlational Concurrent Studies
language (Shore, O'Connell, & Bates, 1984).
Many studies have revealed concurrent associations between In sum, most research shows that children who are more ad-
language development and pretend play (Bates, 1979; Bates, vanced in their play around 1 year of age are more advanced in one
Bretherton, Snyder, Shore, & Volterra, 1980; Casby & Corte, or more aspects of their language around 2. Cross-lagged studies
1987; Corrigan, 1982; Elias & Berk, 2002; Jurkovic, 1978; Lewis, across the second (but not into the third) year are mostly consistent
Boucher, Lupton, & Watson, 2000; Lyytinen, Poikkeus, & Laakso, with the possibility that play could be crucially important to
1997; McCune, 1995; Shore, 1986), with particularly strong cor- language, but the reverse direction of effects is also possible, as is
relations for children under 4 years of age (Doswell, Lewis, Sylva, a third underlying variable. Thus the correlational evidence cannot
& Boucher, 1994).* They have looked at different aspects of distinguish the three accounts; intervention studies are needed.
language (vocabulary size in comprehension and/or production,
syntax) measured in different ways (checklist, free speech, elicited
speech), and different aspects of pretend play (object substitutions, Play-Language Interventions
doll-directed acts, length of play sequences) measured in both free We found four studies that increased play and monitored lan-
and elicited play situations. The evidence that pretend play and guage, three with preschool-aged children. The best control con-
language are related early in development is compelling. Is there dition was in Smilansky's (1968) study training children of immi-
evidence for causality? grants to Israel in pretend play or exposing them to other
"meaningful experiences." Although the play group increased im-
Longitudinal Studies pressively in some aspects of language, statistical analyses were
not conducted.
Longitudinal studies have attempted to determine whether pre- For the other three studies, language increases were seen, but it
tend play predicts symbolic understanding. One study of 10 chil- is not clear what aspect of the intervention contributed. Lovinger
dren tested monthly, from 8 to 24 months,' was conducted along- (1974) engaged 20 low-income 4-year-olds in a daily hour of
side a cross-sectional study of 102 children, six of each age from pretend play for 25 weeks; a control group had no intervention.
8 to 24 months (McCune, 1995). The sample size restricted ana- The experimental group's language increased, but possibly simply
lytic techniques, but McNemar's tests indicated that new levels of due to increased adult contact and conversation. Levy, Schaefer,
pretend play emerged roughly 2 months prior to what was consid- and Phelps (1986) used a pre-posttest design with no control
ered the analogous level of language skill.'" A similar longitudinal group. They expanded a university preschool classroom's pretend
study was carried out with four Japanese children, with similar play area and trained the children in pretend play; in addition, these
results (Ogura, 1991). McCune (1995), like most authors in this children were exposed to more songs, stories, field trips, and
domain, did not assume a causal relationship from play to lan- classroom visitors. Boys' but not girls' PPVT scores increased.
guage, but instead believed an underlying third mechanism was
responsible for the relationship (thus taking the epiphenomenon
view). McCune posited that language emerges later than play ' These studies are not listed in the table because the list is long and they
because it relies on later-maturing vocal control. are not discussed in any detail.
Larger longitudinal samples have permitted cross-lagged corre- ' This is an approximation: For each child, testing was stopped when the
lational and regression analyses. Four studies tested children at the highest levels were achieved.
beginning and end of the second year; only one (Bomstein, Vib- '" Determining what is analogous across the domains of language and
bert. Tal, & O'Donnell, 1992) did not find cross-lagged correla- pretense is a challenge. McCune (1995) sequenced play into five levels and
tions, and it used more limited measures. Two found relationships matched them with verbal ability. Level 1, presymbolic schemes (putting a
between other-directed play early in the second year and measures cup to one's lips), had no lexical match, but Level 2, self-directed pretense
(carrying out the full drinking behavior, perhaps with sound effects or
of comprehension, production, and syntax later in that year (Lyyti-
exaggeration) and Level 3, other-directed pretense (having a doll drink),
nen, Laakso, Poikkeus, & Rita, 1999; Ungerer & Sigman, 1984). were equated with lexical onset (5+ words). Level 4, pretend combinations
Ungerer and Sigman (1984) suggested that other-directed play (pouring before having the doll drink), was equated with multiword onset
indicated an interest in others and communication that is then (producing 3+ word combinations), and Level 5, hierarchical pretense
transferred to language. The fourth study found no association (pretend acts in the absence of perceptual support, e.g., finding a cup for
from play to productive vocabulary or mean length of utterance the doll to drink from), was equated with rule-govemed combinations or
(MLU) but did find a relationship to semantic diversity, or the syntax.
PLAY DEVELOPMENT 19
Table 9
Studies Examining the Effect of Piay on Language
Masked Masked
Type Citation Int Exp Notes
Note. Type of study: C = correlational; CL = cross-lag correlation; E = experimental; T = training. Type of play: SPP = social pretend play; PP =
pretend play (social unspecified). +: positive relationship to play; ~ : no correlation or play = nonplay condition; - = negative relationship to play.
Masking: Intervention (Int) or posttest experimenters (Exp). If masking status was not specified, we assume experimenters were not masked, since that is
the unmarked case. Masking for correlational studies is omitted because it is rarely mentioned, even when it is likely (because play observations occurred
several years earlier than testing, for example). MLU = mean length of utterance.
" Numerous other studies show this; 10 are cited but not described in the section.
The most recent study distdbuted sets of blocks with instructions Narrative
about how to play with them to parents of toddlers (Chdstakis,
Zimmerman, & Gardson, 2007). Low-income children whose fam- Another development that pretend play is claimed to assist is
ilies received this block set, compared with controls who did not, narrative, the ability to tell and comprehend stories. Some have
claimed that pretend play and narrative are different ends of a
had more advanced language postintervention, but (a) we do not
continuum: "Play .. .[is] story in action, just as storytelling is play
know if children pretended with the blocks, and (b) the interven-
put into narrative form" (Paley, 1990, p. 4). If this is the case, then
tion might have increased parent interaction, which then improved
practice or training with either skill should improve the other.
language. More controlled intervention studies are needed to show
Other theodes focus on specific pretense skills that might benefit
whether pretend play specifically might cause language develop-
narrative skills. For example, pretending could foster metalinguis-
ment.
tic skills necessary for storytelling, because dudng social pretend
play communication often occurs at the meta level (Garvey &
Literacy Kramer, 1989; Giffin, 1984). In addition, pretense role play re-
Ample research has shown that exposure to literacy play mate- quires a child to imagine and track the perspective of another
dals (like plastic letters or a model post office) increases literacy character, a skill also used when following and telling a story
(Neuman & Roskos, 1992; Roskos & Neuman, 1998). This is (O'Neill & Shultis, 2007; Ziegler, Mitchell, & Curde, 2005).
similar to findings with board games and math (Ramani & Siegler, Finally, embodied cognition (Lillard, 2005, Chapter 2; Scott, Har-
ds, & Rothe, 2001) would predict children would remember sto-
2008). Although it is useful to know that one can influence what
des better after acting them out. We located 14 studies of whether
children play with, and through those play matedals can influence
pretend play leads to better narrative skills, operationalized as
their skills, it is particular to specific pretend play content and not
storytelling, memory, and comprehension (see Table 10).
pretend play generally. Increased exposure to that same content
outside of pretend contexts might be equally effective. What could
be important here with regard to pretend play is motivation. If Correlational Studies
pretend play is a context that especially motivates engagement
with literacy matedals, this is significant. Three studies examined whether children who naturally en-
gage in more pretend play have better narrative skills. In the
initial one, Johnson (1976) coded the free play of 3- to 5-year-
Summary
olds as nonfantasy, social fantasy, or nonsocial fantasy and later
Children's levels of pretend play and their early language de- gave children a story completion task. Scores were significantly
velopment do appear to be related, with pretending preceding correlated with social but not solitary pretend play. Trionfi and
language. Researchers in this area tend to think that the domains Reese (2009) compared 5-year-olds with and without imaginary
are related due to an underlying symbolic function—an epiphe- companions (ICs), and children with ICs scored higher on
nomenal reason. A causal account is possible, although a reverse narrative quality (but not memory) when retelling a story and
direction of effects is as well, and better intervention studies are telling personal narratives. Both of these studies suggest a
needed to determine which of the three models is best supported in possible causal relationship from pretend play to narrative. A
this domain. longitudinal correlational study would provide better evidence.
20 LILLARD ET AL.
Table 10
Studies Examining the Effect of Play on Narrative Skills
Note. Type of study: C = correlational; CL = cross-lag correlation; E = experimental; T = training. Type of play: Solo = pretend play alone; SPP =
social pretend play; D = enacting stories with dolls or other children; IC = imaginary companion. -I-: positive relationship to play; ~: no correlation or
play = nonplay condition; - = negative relationship to play. Masking: Intervention (Int) or posttest experimenters (Exp). If masking status was not
specified, we assume experimenters were not masked, since that is the unmarked case. Masking for correlational studies is omitted because it is rarely
mentioned, even when it is likely (because play observations occurred several years earlier than testing, for example). Tr = teacher.
In the third correlational study researchers coded the social Benson found that children in the control condition actually told
pretend play talk of 48 Head Start children and their mothers more complex stories. The third study involved retelling a story
during a free play session in both kindergarten and fourth grade that had been read to them rather than generating an original
(Tenenbaum, Snow, Roach, & Kurland, 2005), and when the story (Kim, 1999), as well as answering questions about it
children were in sixth grade, told them a novel story and then (considered below). Children given toy animals produced more
tested for comprehension. Controlling for maternal education complex and complete narrative retellings than children who
and age 3 verbal ability, neither child's nor mother's pretend were given pictures of the same animals. Results are thus
play talk in kindergarten or fourth grade predicted sixth grade inconsistent as to whether pretense-supporting props lead chil-
story comprehension. Thus narrative production but not com- dren to tell better stories; perhaps pictures (as in Benson's
prehension or memory might be associated with pretend play. control condition) help children more with making up new
One possibility here is that having a high fantasy predisposition
stories, and props are more helpful when the task is to retell a
results in (a) pretending a lot with one's peers, (b) having an IC,
previously heard story. Props might assist story retellings via
and (c) telling elaborate stories; this is an epiphenomenal ex-
embodied cognition.
planation. However, experimental studies are the best source of
Supporting this possibility are three other studies looking at the
evidence for whether and how pretend play affects narrative
skills. impact of pretense reenactment on story memory. Pellegrini and
Galda (1982) compared three groups of 5- to 7-year-olds, each of
which had two training sessions followed by a test session. In all
Experimental Studies three sessions a story was read, and then a pretend play group
Three studies have addressed whether children tell better reenacted the story, a discussion group answered questions about
stories when they are pretending, with pretending operational- it, and a drawing group drew pictures about it. Children in the
ized as being provided with small model toys. Ilgaz and Aksu- reenactment group scored better both at retelling the story and
Koç (2005) and Benson (1993) contrasted the stories children answering questions about it. Pellegrini (1984) also found that
told with and without toys. The former study had no external peer-directed play was as effective for story memory as adult-
supports for the control condition, whereas Benson provided the directed play.
control group with drawings of characters. Ilgaz and Aksu-Koç At issue is whether the pretend play was the source of this effect
found no condition differences on a first pass. They then versus embodied cognition (moving one's body in ways that
excluded the youngest few children in each age group (3-, 4-, represent the story). Testing this, Marbach and Yawkey (1980)
and 5-year-olds, original n = 10 per age, within-subject design) compared 5-year-olds' recall of a story in two pretend conditions:
and conducted another analysis of variance with 20 children. reenactment with one's whole body or with a puppet, thus only
With this approach, they obtained a near-significant result (p = with one's hands. Although both experimental conditions could be
.05) of more complex stories with toy props for the older considered pretend play, children who reenacted the story with
4-year-olds only. Especially with the decision to exclude certain their whole body showed significantly better recall than children
subjects being post hoc, this is properly considered a null result. who did so with a puppet. Kim (1999) also supported this: Chil-
PLAY DEVELOPMENT 21
dren who retold the story with toy props were no better at answer- Another study used many more teachers—13 in all (Silvern,
ing memory questions about the story than were ehildren who had 1986). Children from kindergarten to third grade (n = 505) aeted
pietures of the animals. out or were simply read stories, preeeded and followed by 10
In sum, research suggests that pretend role play, but not toy multiple-ehoiee questions. Teachers administered all aspects of the
props/puppets, leads to better memory for stories, due to embodied procedure and served as their own eontrols by administering the
eognition; also when retelling but not when making up new stories, experimental eondition to their own class and the eontrol eondition
having toy props leads to more elaborate narratives. Thus eertain to another one. Although on face this seems like a good idea,
types of pretend play influence eertain aspects of narrative devel- teaehers have a personal stake in their own class (the experimental
opment, but the effects are limited. one) doing well; having multiple disinterested outsiders as exper-
imenters would be better. Pretend training did result in signifi-
Training Studies cantly more gain for teachers' own classrooms. Further analysis
showed this was only for ehildren with poor narrative skills (Wil-
We located six studies that trained ehildren in pretend play and liamson & Silvern, 1990). Although narrative skill and IQ might
then examined narrative skills. Saltz and Johnson (1974) had be orthogonal, the eontrast with the Saltz et al. (1977) finding that
preschoolers act out stories (thematic fantasy play) or label and only higher IQ ehildren were affected by training is worth noting.
classify objeets. A narrative memory test involved arranging six Exeept for the soeiodramatie play eondition in Saltz et al.
pietures in proper sequence while retelling a story, and a story (1977), the intervention studies reviewed so far have involved
generation task involved making up new stories with sets of five reenaeting stories. Dansky (1980a) used more everyday pretend
pietures. Thematie fantasy play led to significantly higher scores play, with low-income preschoolers in three 30-min sessions per
on the story memory task and longer stories with more inferences week for 3 weeks. In the soeiodramatie play group, 12 ehildren
and eonnectives in the story generation task. Although initially
were encouraged to play out everyday themes. A free play group
compelling, this study has several problems, such as lack of
received no direction in their play, and an exploration training
masked experimenters and teaching to the test. On the latter point,
group discussed objeets in a manner that appeared to equalize adult
fantasy training involved going over stories repeatedly including
eontact. The intervention increased the amount of pretend play in
answering questions about them ("Why did the billy goats eross
the first group, whereas the other groups—like the first group
over the bridge?"), so one cannot know if it was the acting out or
preintervention—showed little pretend play. Masked experiment-
simply the deeper consideration of stories that helped.
ers administered three narrative posttests of story memory and
Saltz et al. (1977) addressed teaching to the test by including a quality, and children in the soeiodramatie training group seored
fantasy story diseussion group along with fantasy play training; the higher on nine of 10 measures. This study is promising for the
other two groups were soeiodramatic play training and an arts and hypothesis that pretend play eauses narrative skills, espeeially
crafts control. On two narrative tasks similar to the ones used because the methods were sound (the free play group had less adult
earlier, all ehildren did poorly, and there were no eondition dif- contact, but eontaet appears similar for the exploration group) and
ferenees. However, for the story generation task and the high IQ ehildren were not acting out stories but rather were merely eneour-
subset of children, a signifieant difference was found between the aged to engage in the types of everyday pretend they might engage
two pretend play and two nonpretend play conditions. Yet even for in on their own. Replieation with a larger sample size would be
this subset, scores were between 0 (naming pictures but with no
helpful.
relation to others) and 1 (minimal story, no elaboration). Thus an
unplanned analysis led to a small result on one of several measures
with unmasked experimenters for a subset of children. Summary
In a more tightly controlled study that took the precaution of
using masked experimenters, Baumer, Ferholt, and Leeusay (2005) The research reviewed here suggests that providing toys does
tested the effeets of a kindergarten pretend play intervention im- not enhance new stories but does help with story retelling. Story
plemented for 2 hr/week for 14 weeks. In one experimental and memory is helped by role play, probably due to embodied eogni-
one control classroom, the intervention began with hearing and tion. Children who pretend more also appear to tell more elaborate
diseussing The Lion, the Witch, and the Wardrobe. Then in the stories, although when older they were not better at story eompre-
pretend play classroom, four experimenters reenacted passages hension. Thus experimental and eorrelational studies had qualified
from the book, and eventually the teaeher and then the children results for the hypothesis that pretend play eauses narrative devel-
joined in. In the control classroom, the teacher continued to read opment. Although most of the training studies have methodolog-
and discuss the story, and the same experimenters drew pictures ieal shortcomings, the strongest one suggests that social pretend
and wrote stories with the ehildren. Results showed that ehildren in play positively influences narrative development. This would
the pretend play classroom had significantly higher posttest scores make sense, since soeial pretend play does involve creating and
for story comprehension and story length and coherence (but not acting out narrative. Still, this evidenee does not make a solid case
eomplexity). Three concerns with this study are small samples (12 that pretend play is crucial for narrative development. Further
eontrol, 17 experimental), classrooms with very different ethnie researeh should examine additional alternate routes, like story
compositions that eould influenee narrative trajectories, and the reading without enaeUnent. Although Baumer et al. did this, the
use of one full-time classroom teaeher per eondition—the teachers eonfounding of intervention and elassroom teacher is problematic.
might have differed in narrative skill teaching effectiveness out- In addition, researehers in this area should be eareful to distinguish
side the intervention. Although the results are intriguing, replica- the different aspects of narrative development under consideration;
tion is needed. telling stories, story memory, and story comprehension.
22 LILLARD ET AL.
Pretend Play and Self-Regulation comes to pretend play specifically, however, results conceming the
relationship to EF ate inconsistent.
Pretend play is sometimes claimed to improve childten's self- Two older studies used the ability to stand still as an index of
regulation (Betgen, 2002; Bredekamp, 2004), which involves the EF. J. L. Singer (1961) rated 6- to 9-yeat-olds as high ot low in
top-down cognitive processes called the executive functions, and fantasy, based on fantastical toy atid game preferences and having
(possibly overlapping) processes that regulate one's level of emo- an imaginary companion, then asked them to remain as still as
tional arousal (Blair & Raver, 2012). We focus on each in tum. possible for up to 15 min. The covet story justifying this request
was that the researchers were conducting "a study of persons
Executive Function suitable for space flight with its long pedods of solitary confine-
ment in a narrow space [to find] space men of tbe future'" (p. 404).
Executive function (EF) is an umbtella term fot a suite of related
High fantasy children temained still twice as long as low fantasy
skills including inhibitory control, working memory, and attention
childten. Children's higb imagination appeared to help: Some of
(Blair & Razza, 2007; Gaton, Bryson, & Smith, 2008; Miyake,
the children who waited the longest told tbe expedmentet, in
Fdedman, Emerson, Witzki, & Howettet, 2000; Rueda, Posner, &
Rothbatt, 2005). Vygotsky (1978) claimed pretending helps such posttest interviews, that they had been engaged in fantasy while
skills based on two aspects of pretend play. Fitst, it involves waiting (e.g., "playing a tocket game," p. 408).
behaviors that ate internally guided as opposed to stimulus bound. Manuilenko (1975) examined conditions undet which 3- to
Pretending involves operating at two levels, the real and tbe 6-yeat-olds could stand still the longest. Two conditions embedded
pretend (Lillard, 1993); the pretend level is largely internally tbe instruction in a pretense game. Factory and Guards, in wbicb
guided, while the real (external world) level must in many ways be tbe guards were to stand motionless while other children did tbe
inhibited. In this interpretation all pretending should be related to factory work (packing boxes). Four-year-olds appeated to be most
EF. Second, Vygotsky noted that pretend role play involves be- successful (no statistical analyses were done) when standing guard
having according to norms that often differ from one's own ev- ovet two to four cbildren packing boxes. They were less successful
eryday behaviors. Studies of ptetend play and EF (see Table 11) when standing guard outside tbe classroom, out of sight of tbe
bave used teacher or parent ratings of self-regulation and direct other players or for other, nonpretense reasons. Younger and older
tests of EF. Some have coded naturalistic pretend play, and others children were not as affected by the pretend scenado.
have used laboratory measures. Taken together, these two older studies suggest tbat specific
Correlational and experimental studies. Several studies ptetend play content can motivate and strengthen voluntary behav-
have reported that children with better self-regulation are more ior for some children. Tbe limitations (high fantasy cbildten, one
likely to engage in positive peer play generally (Fantuzzo, Sekino, of four age groups, social context) are problematic for tbe idea that
& Cohen, 2004; Méndez & Fogle, 2002; Mikami,.2010). When it pretending causes EF generally. This position would be better
Table 11
Studies Examining the Effect of Play on Self-Regulation (Executive Eunction and Emotion Regulation)
Masked Masked
Type Citation Int Exp Notes
Manuilenko (1948/75) PP No No
E Child interview -I-; play observation ~ ;
Cemore & Herwig (2005) PP
C teacher interview ~ ; parent interview ~
c Albertson & Shore (2009) PP Lab measure -I-; did not examine own
naturalistic pretense
c Carlson et al. (2012) PP PP Lab measure +; own naturalistic pretense ~
c Kelly et al. (2011) PP PP Lab measure -I-; own naturalistic pretense ~
c Elias & Berk (2002) PP PP Limitations: only clean up measure, and only
in high-impulsive children
c Harris & Berk (2003) PP Head Start children
T Saltz et al. (1977) PP PP No No Failed to replicate from one year to next
T Diamond et al. (2007) PP No No Unclear if play or other aspects of intervention
C/E Galyer & Evans (2001) PP PP Yes Yes Mixed results in experiment; also found
correlation
E L. A. Bamett & Storm (1981) PP No No Unclear if play or time
E L. A. Bamett (1984) PP No No Unclear if due to negative impact of control
condition or positive impact of play
T Moore & Russ (2008) PP No No
Note. Type of study: C = correlational; E = experimental; T = training. Type of play: PP = pretend play (social unspecified). +: positive relationship
to play; ~ : no correlation or play = nonplay condition; - = negative relationship to play. Masking: Intervention (Int) or posttest experimenters (Exp).
If masking status was not specified, we assume experimenters were not masked, since that is the unmarked case. Masking for correlational studies is omitted
because it Is rarely mentioned, even when it is likely (because play observations occurred several years earlier than testing, for example). EF = executive
function.
PLAY DEVELOPMENT 23
supported if children who pretend a lot show higher EF outside of have rarely been shown, and even then appeared only for one of
the play situation. two measures, and only with high-impulsive middle-class children,
Cemore and Herwig (2005) tested this with a standard delay not low impulsive or lower class ones. This pattem of results does
of gratification test (with no cover story) and by assessing play not support that pretense is key to developing EF. But as in other
in three interviews (mother, teacher, child) as well as via direct domains, training studies provide the best test.
preschool observation. Only the child interview was related to Training studies. We found three studies examining the
delay time, but this interview occurred just after the delay task effect of training pretend play on EF. In a study discussed in
with the same experimenter, leaving open the possibility of several earlier sections, Saltz et al. (1977) used thematic fantasy
experimenter biasing. None of the other measures of play was play, sociodramatic play, fantasy discussion, and controls. Recall
significantly related to EF. that this study used unmasked experimenters and subsets of data
Three other studies found a relationship to laboratory tests of without prior rationale. Here there is rationale, because although
knowledge about pretend play (Albertson & Shore, 2009; Carlson, the study involved three cohorts, EF tasks were given only to the
White, & Davis-Unger, 2012; Kelly, Hammond, Dissanayake, & first and third. First year children were "guardians of the toy":
Ihsen, 2011); however, the knowledge tests themselves appear to They had to sit by an attractive toy and if they did not touch it
require EF. What would be most telling is if engaging in pretend during the waiting period, they were later rewarded with another
play were itself related to EF. The latter two studies measured toy. Two thirds of the children were also given instructions about
spontaneous pretend play and found no relationship to it. how to make waiting easier (Mischel, Shoda, & Rodriguez, 1989):
Perhaps one can see a relationship from pretend play to EF only "think about your favorite story" or "read this book" (p. 371).
over time. In a cross-lagged correlational study, Elias and Berk Children in the play training conditions (combined) waited longer
(2002) examined fifty-one 3- to 4-year-olds in four classrooms at than children in the other two conditions; although an interaction
two times points 4 months apart, recording play behaviors at the with instmctions was not significant, the authors stated that "the
first time point and self-regulatory behaviors (attentiveness at data clearly display an interaction" (p. 376), limiting the result to
circle time and behavior at cleanup time) at both time points. the subset of children instructed in how to make waiting easier.
Pretend play was coded as solitary, social, and complex (e.g., This result did not replicate in the third year, however, when the
adopting a role, talking in that role, and using substitute objects). task was sitting in a chair for 5 min pretending to be in a spaceship.
Partialing out age, verbal ability, and Time 1 cleanup or circle time Two other studies are cited as showing that pretend play assists
attention, complex social pretend play predicted Time 2 cleanup, EF (Hirsh-Pasek et al., 2009; Nisbett, 2009), although we question
but not Time 2 circle time attention. Further analyses showed the whether pretend play is the source of the effect. Tools of the Mind
cleanup relationship existed only for children rated high in impul- is a Vygotsky-inspired preschool program with many "tools" to
sivity; pretend play made no difference for low-impulsive children. assist self-regulation. For example, one tool is a freeze game, in
The authors' explanation for the lack of relationship to the circle time which children run around to music, and when the music stops all
measure is that the teacher more closely supervised circle time. children must freeze until it starts again, directly challenging their
However, children were not at ceiling on this measure at Time self-regulation. Another is "reading buddies," in which one child
2—^they scored an average of 104 out of a possible 120, with a holds a symbol of an ear while another holds a symbol of a mouth,
standard deviation of 16.8—so it is not clear that closer teacher to help them remember if it is their tum to listen or to speak.
supervision explains the lack of association. Regarding pretending, the program has an unusual requirement:
Most conceming about this study, however, is failure to repli- Before they begin play, children must draw or dictate to the teacher
cate the result with low-income children (Harris & Berk, 2003, as a play plan; once play commences, they are not allowed to deviate
cited in Berk, Mann, & Ogan, 2006). In a similarly designed study. from that plan (Diamond & Lee, 2011). Of five studies that have
Head Start children's Time 1 sociodramatic pretend play was looked at the EF outcomes of this program, one had strong results
actually negatively related to their Time 2 cleanup behavior (r = (Diamond, Bamett, Thomas, & Munro, 2007), one had milder ones
-.25). If social pretend play helps self-regulation, one would (W. Bamett et al., 2008), and three others saw no impact (Cle-
expect it to do. so for low-income children as well—indeed the ments, Sarama, Unlu, & Layzer, 2012; Lonigan & Phillips, 2012;
Fantuzzo et al. (2004) positive correlation with social play was S. J. Wilson & Farran, 2012). Even were results consistentiy
found for Head Start children. Berk and colleagues explained the positive, one cannot separate pretend play from other aspects of the
discrepant results as being due to the themes of children's pre- program, preventing a clear conclusion that pretend play influ-
tending, which they said were often aggressive in the Head Start ences EF. Other aspects of this multifaceted program might be the
group. If that is the case, then one cannot say that pretend play engines.
helps self-regulation generally (and as the Elias & Berk, 2002, In sum, training studies of pretending and EF are entirely
study is often cited as showing, e.g., Bredekamp, 2004; White- inconclusive, with failures to replicate and uncertain causes. Cou-
bread, Bingham, Grau, Pino Pastemak, & Sangster, 2007); rather, pled with very limited correlational findings, evidence that pretend
complex social pretend play with particular content (Elias & Berk play assists EF is sparse at best.
stated the main play theme was "housekeeping") might help more
impulsive children's self-regulation as assessed by a cleanup mea- Emotion Regulation
sure but not an attention measure.
In sum, strong evidence that pretend play helps EF is lacking. Others have focused on pretense assisting emotion regulation.
Two older studies showed that when pretending a role where Freud (1955) saw all play as releasing tension, and Erikson (1950)
standing still is important, pretend play helps some children under believed children could master disturbing events through play.
some eircumstances. Relationships to spontaneous pretend play Fein (1989) and Bretherton (1989) went so far as to claim emotion
24 LILLARD ET AL.
regulation is the pdmary function of pretend play. This possible These two studies suggest that solitary pretend play might
function of play relates closely to pretend play therapy. For reasons reduce anxiety. But in the first study, the passage of time could be
of space we do not review this literature here but note that a current the crucial factor, and in the second study, the reduction could be
review from within the field, which considered two recent meta- only relative to a rather restdctive control activity. Further research
analyses, concluded that evidence for play therapy's efficacy is is needed to examine these possibilities.
"largely inadequate" (p. 13) and that the strongest effects are seen A training study, mentioned earlier regarding creativity, found
in studies of children who play out theme-specific issues concern- no immediate nor long-term effect of pretend play on teacher-rated
ing medical treatment (Phillips, 2010). Thus, pretending about emotion regulation (Moore & Russ, 2008). Forty-five first and
specific content might help children cope with that content in real second grade children were trained in pretend play and 2 to 8
life, but in general we lack good evidence for the efficacy of months later were tested. Although their pretending had increased
pretend play therapy. relative to that of controls, and so had their positive affect expres-
We found four studies focused on pretend play and emotion sion, the trained children did not improve in emotion regulation as
regulation. In a correlational study, parents of 4-year-olds filled measured by the teacher. Perhaps positive affect expression re-
out a measure of emotion regulation (Shields & Cicchetti, 1997) flected an improvement in underlying emotion regulation that the
and reported on their child's play at home (Galyer & Evans, 2001), teachers had as yet failed to detect, a speculation that requires
and they found that children who were rated as higher in emotion further research.
regulation were also reported to pretend more at home. Then a Taken together these four studies leave open the case as to
laboratory measure was given to determine if children who have whether pretend play assists emotion regulation. Although parents
higher emotion regulation are better pretenders. Children were rate children similarly on both, expedmental paradigms have im-
engaged in disrupted pretend play: In the middle of a nice episode, portant altemate explanations that need to be ruled out, and the
a toy crocodile suddenly threatened to eat all the other toys. The single training study we found failed to find that pretend play
parent emotion regulation scores were not significantly related to training increased emotion regulation as measured by the teacher.
how long the children spent in pretend play with the researcher or
how effectively the disruption was handled, but they were more Summary
likely to continue the play. This reflected only a 2-point difference
in emotion regulation score (26 vs. 28 of 60) for children who did Some scholars have claimed that pretend play improves, and
and did not continue the play; further, the emotion regulation even is crucial to the development of, self-regulation (Bergen,
measure was not validated on 4-year-olds. The small difference 2002; Bredekamp, 2004; Hirsh-Pasek et al., 2009; Tomlinson,
might not be of practical importance, although the naturalistic 2009). This review shows little support for such claims. The
correlational result is interesting. inconsistency in correlational studies is against a general causal
In a study in which specific pretense content appears to be key, account. Without further research, there is no basis to determine if
L. A. Bamett and Storm (1981) showed forty 3- to 5-year-old equifinality is supported; the direct path might not exist, and no
children a traumatic episode of the television show Lassie, and half other paths have been examined. Epiphenomenalism adses when
the children (the "unresolved" group) did not see the episode's there are inconsistent correlations, but in this literature correlations
happy resolution. Afterward, those in the unresolved group chose are so sparse there might be nothing to explain.
to play significantly more with a toy Lassie, and their play more
frequently involved themes fi"om the episode (although they did Summary and Conclusions
not reenact it). Their anxiety (Palmar Sweat Index) also was
significantly reduced after the play, and their self-reported happi- This review has examined evidence cited to support claims that
ness was increased. Since there was no control that saw the pretend play is a crucial engine of child development. In addition
unresolved episode but did not play, it is not clear how much the to the causal Vygotskian model inherent in such claims, we con-
passage of time versus the play itself is responsible, but the finding sidered two altemative possibilities: that pretending is one of
that children chose to play with Lassie and then their anxiety was several possible routes to development, or that pretending is
reduced is interesting. merely an epiphenomenon, something that often goes along with
A later expedmental study examined pretending's impact on important developments, but does not cause them. The overdding
stress following mother's departure on the first day of preschool conclusion from this review is that there is currently not evidence
(L. A. Bamett, 1984). Seventy-four children (mean age 3.3 years) to support the first position and that more and better research is
were divided into eight conditions. Roughly half of the children in needed to cladfy pretend play's possible role in children's devel-
each group were not very anxious at their mother's departure and opment. Table 12 summadzes findings from this review as regards
half were (Palmar Sweat Index). Within each of these groups, half the three possible models, which will now be discussed in tum.
were told to sit at a table and to hear a story about local vegetation, The causal position is that pretend play has a unique and
and half were brought to a room full of toys to play freely. Within important role in promoting healthy development. This might seem
each of these two conditions, half of the children were alone, and like a straw-person view, but the claim is repeatedly made in the
half were with five other children. Thus the conditions were literature, as we have shown throughout this article. If this position
social/alone nested in story/play nested in high/low anxious. For were supported, then for any development pretend play causes,
the high anxious children, playing with toys alleviated anxiety strong, consistent, and unique correlations should be seen between
more than listening to the story. In addition, playing alone allevi- pretend play and the development. We concluded that the causal
ated anxiety more than social play. Low anxious children showed account is possible, based on existing research, for four of 11
little effect of condition. developments reviewed here: reasoning, language, narrative, and
PLAY DEVELOPMENT 25
Table 12
Summary of Evidence for Each Position
Domain or
subdomain Causal Equifinal Epiphenomenal
Creativity No: Inconsistent correlations. No: When experimenters are Best supported, but not clear what of.
masked or filmed or have Adult interaction, materials, social
other hypotheses, null mix?
results.
Intelligence No: Although correlations, direction No: Music training is more Best supported: Adult interaction or other
of effects is uncertain; skills effective. features of intervention.
training suggests adult interaction
could be underlying third
variable.
Problem solving No: Construction but not pretend No: Construction but not No: Construction but not pretend play.
play. pretend play. Associations might result from
propensity to construct.
Reasoning No: Getting children to focus on Yes. Possible: If pretend as operationalized is
premises is as effective. also a cue to pay attention to premises,
and this is true reason for results.
Conservation No: Correlational studies find no No: When experimenters are Best supported: Adult interaction
relationship; training results ride masked and other aspects (structured questioning).
on adult questioning. of intervention equalized,
null results.
Theory of mind No: Inconsistent. Some correlations Possible, yet sounder Best supported, considering inconsistent
to social pretend play with more methods fail to show. findings and hints of reverse direction
recent tasks, but direction of of effects. Adult interaction.
effects is unclear.
Social skills No: Correlations inconsistent with Possible: Other routes Possible: Crucial variable could be
both solitary and social pretend unexamined. practice.
play. Direction of effects is an
issue.
Language Possible: Consistent relationships to Possible: Other routes Possible: Adult interaction could explain
different aspects of language. unexamined. training study results.
Effects could be bidirectional.
Narrative Possible: Correlations inconsistent Possible. Less likely but one solid but small
and to different aspects of training study needs replication.
narrative development but one
solid but small training study
needs replication.
Executive function Not likely: If so, limited to subsets Not clear that pretend play Not clear that pretend play is reliably
of children and tests. leads to. associated.
Emotion regulation Possible: Parent rating consistent in Not clear that pretend play Too few studies.
single study; other results have leads to.
other interpretations.
emotion regulation. Of these developments, the causal account is (intelligence), or results were limited to subsets of children or
most plausible for language, as pretend play is quite consistently measures (executive function).
related to it. However, correlation is not causation, and reverse Equifinality was supported for the domain of reasoning, since
causality (from language to play) is shown in some studies. An studies have clearly shown that pretend play is one way to get
underlying variable like adult interaction could also be important, children to focus on the premises and perform better on logical
with pretend play possibly being epiphenomenal to intensive de- syllogism tasks. However, one might also argue that this supports
velopmentally oriented adult interaction, explaining results from epiphenomenalism: The pretend play aspect of the experimental
training studies. For narrative development, conclusions were sim- condition was epiphenomenal to the true underlying cause of
ilar, although the database is not as dense and the domain is focusing on the premises.
complicated by the different aspects of narrative showing some- Equifinality is also possible for some other domains, although
what different results (unlike language, where results were fairly few other routes have been explicitly compared—this is a topic for
consistent regardless of aspect of language measured). Emotion further research on the development of social skills (music training
regulation and reasoning are more difficult to evaluate given the was not as good), language, and narrative (free play was not as
scarcity of solid research. For all other areas, the causal account is good, story discussion route requires further study). For theory of
not supported by available research: Correlations to pretend play mind, in the most solid study neither pretend play training nor
were inconsistent for no clear reason (creativity, theory of mind, skills training led to significant improvements; other studies had at
social skills) or did not exist (conservation, problem solving), or least one serious methodological problem. For creativity and con-
pretend play had no unique role since other training worked as well servation, equifinality was not supported because masked experi-
26 LILLARD ET AL.
menters eliminated the eausal result. For intelligence, equifinality ments that would then help on the outeome test, eould plausibly
was not supported beeause musie lessons led to greater gains than eause results. If ehildren's pretending involves assigning props to
drama lessons. For the two aspects of self-regulation, there are not roles, they do better on ToM tasks that involve assigning objeets to
elear findings showing that pretend play is a possible route; exist- people, and if they pretend by focusing on the narrative struetures
ing findings are too sparse, limited, or attributable to other faetors. of stories, their own stories come to have better narrative strueture.
Epiphenomenalism is less likely for narrative beeause of the one This is important in that ehildren are naturally motivated to play,
small but solid training study (Dansky, 1980a), but it eould explain and if we ean embed leaming in play materials sueh that we
some findings. For example, high fantasy ehildren might produce positively influenee development, this eould be good. It is the
more complex narratives and also be apt to pretend more and have same rationale underlying many technology toys, and even Sesame
ICs. For self-regulation, there are too few findings showing a Street. But just as watehing television generally does not help
relationship. Epiphenomenalism seemed possible for language, development even though watehing partieular eontent ean, the
social skills, reasoning, and problem solving; and for creativity, evidenee reviewed here suggests that pretend play tnight not
intelligenee, eonservation, and theory of mind, we believe it is the generally help development on its own but that playing with
best supported position. particular eontent ean. It is notable that in one study, the more that
Two truths are abundantly elear firom this review. One is that we "hard to manage" ehildren pretended, the worse off they were
do not have a good basis of evidenee from whieh to elaim that developmentally—but their pretend eontent was often violent (J.
pretend play is crucial to development. The seeond is that much of Dunn & Hughes, 2001). In sum, features of the adults with whom
the evidenee on pretend play suffers from serious methodologieal ehildren interact, features of the children themselves, and the
problems that must be addressed in further researeh to allow for a eontent with whieh ehildren pretend are potential epiphenomenal
solid assessment of whether pretend play eauses any important reasons for some findings relating pretend play to positive devel-
development. But as a thought experiment, we next take seriously opmental outeomes.
the third position, asking of what pretend play might at least
sometimes be an epiphenomenon.
Implications for Educational Settings
Three Epiphenomenal Reasons for Some Findings Despite the poor state of the evidenee on pretend play's benefits,
researeh does not advocate what is often offered as the only altema-
If pretend play is at least in some domains epiphenomenal, then tive to a playful approach in educational settings: adult-eentered
of what might it be an epiphenomenon? It is not likely to be just instruction. Research in U.S. sehools has clearly shown that adult-
one thing, but we hypothesized in several eases that eorrelations eentered leaming environments are less positive for young ehildren
between pretend play and positive developmental outcomes eould than more aetive, ehild-eentered approaehes (Stipek, Feiler, Daniels,
be due to the same adults influencing children's development in & Milbum, 1995) dubbed "playful learning" (D. G. Singer, Golinkoff,
both domains. Adults who hear the child development rhetorie & Hirsh-Pasek, 2006), like Reggio Emilia, Montessori, and Tools of
noted at the beginning of this artiele and have time and interest to the Mind. Developmental science does not support young ehüdren
devote to their children's psyehologieal development would en- sitting in desks while teachers lecture at them.
eourage pretending, along with other development-enhancing ac- What else about ehild-eentered classrooms leads to more posi-
tivities. When adults eneourage pretending, ehildren pretend more tive developmental outeomes if it tums out not to be the pretend
(Lillard, 2011). But where pretending is not eonsidered so impor- play? Child-eentered classrooms differ from teacher-eentered ones
tant, ehildren learn from adults in other eontexts (Berk et al., in several qualities. Like pretend play, ehild-eentered elassrooms
2006), and then correlations between pretend play and the outcome often provide free choice, interesting hands-on activities for whieh
do not exist. To examine this hypothesis will require more careful the ehild is intrinsically motivated, and peer interactions. Unlike
study of the relation between pretense and other developments in pretend play, these elements have been shown in independent
the eontext of adult-ehild interactions. researeh to be consistently assoeiated with more positive outeomes
A seeond possible epiphenomenal reason for positive results (see summaries of the literatures in Lillard, 2005). Compared with
eoneems features of ehildren that would also go along with higher free play programs, more struetured elassrooms with earefully
pretend play seores in the studies. In some eases, we noted that designed, challenging, hands-on activities that eonfer leaming
ehildren who are more soeially compliant and/or intelligent would appear to help children's development the most (Chien et al., 2010;
do better on pretend play and the other tasks, sinee the researcher Lillard, 2012; Lillard & Else-Quest, 2006).
was asking them to pretend and asking them to engage in some Do these findings regarding pretend play mean ehildren need no
other task. For language, a plausible altemative ehild factor is a time to play (pretend or otherwise)? First, there is good researeh
"symbolic funetion" supporting both language and pretend play. showing that recess restores attention in eonventional school set-
Indeed, researeh has shown that children's performanee on sym- tings where the basie instruetional method involves ehildren sitting
bolie tasks that are not primarily eoneemed with language or at desks and listening to teaehers (Pellegrini & Smith, 1993). In
pretend play (DeLoache, 2000) is assoeiated with both (Ka- addition, exercise, be it from athleties or reeess, improves eogni-
vanaugh & Lillard, 2012; Walker & Murachver, 2012). tive function (Lillard & Erisir, 2011; Ratey, 2008). A perfeetly
A third underlying faetor that eould explain some positive sufficient reason for play time might simply be that it is fun
experimental results is the eontent about whieh ehildren are asked (Power, 2000). Pretend play is also relaxing, assoeiated with more
to pretend. In studies of problem solving, theory of mind, execu- heart rate variability (Hutt et al., 1989, p. 12). Finally, the researeh
tive funetion, and narrative, teaehing to the test, or using eontent in reviewed here often suggested that adult interaetion might be the
the pretend training or manipulation that directly involves ele- real underlying eause of positive effeets from various interven-
PLAY DEVELOPMENT 27
tions. Ptetend play migbt be useftil because it is a setting tbat can pottant. Tbe Test of Ptetend Play (Lewis & Boucber, 1997) ad-
facilitate positive adult-cbild interaction (Paley, 2005; P. K. Smitb dresses levels of play dedved from observational studies.
et al., 1981). Analyses. A recurrent problem witb tbe existing reseatcb is
bow results wete analyzed and reported. Btainetd (1982) pointed
Implications for Methods out tbat even when positive effects of ptetend play were obtained,
tbey wete tiny. Expedmentets in tbis area bave not always been
Tbe literature on tbe possible benefits of pretend play for de- dgorous in tbeit analyses, perbaps because of tbe play etbos (P. K.
velopment sbowcases an array of methodological problems. Fut- Smitb, 1988). In some instances, no result was found on an
tber research in tbis area must avoid these problems to elucidate if omnibus test but follow-up tests were done. One-tailed tests were
and how ptetend play might help development. Here we highlight used without strong rationale. Only subsamples were included, for
a few major problems that must be addressed (see also Cbeyne, no compelling reason. Unfavorable tesults were ignored. Tbe
1982; Cbdstie & Johnsen, 1983). existing meta-analysis (E. P. Fisher, 1992) used ettoneous and
Experimenters. In the literature on pretend play, expedment- cbetty-picked statistics. Solid trutbs stand up to dgorous analyses,
ets bave rarely been masked, but when they wete, results often and undgorous tecbniques only muddy tbe waters.
went away, suggesting experimenter bias created the odginal re- Other recommendations. Altbougb our own ptefetence is
sult. It is possible tbat masked expedmenters sometimes result in expedmental metbods, otber researchers favor correlational de-
null effects because cbildren are not fatniliat witb tbe second signs witb natural settings. Modem statistical tecbniques allow
expedmenter. If tbis is tbe case, tben tbe positive influence of causal inferences from such designs witb sufficiently large sam-
pretending is so tenuous tbat being tested by an unfamiliar posttest ples and numerous measures. Since ptetend play migbt have ef-
expedmenter dwarfs it. We tbink tbis unlikely (Lillatd & Peterson, fects only ovet a long time coutse, longitudinal efforts sbould be
2011). More likely it seems tbat expedmentets who ate swayed by encoutaged.
tbe cultutal view of play and knowledgeable about condition Finally, we would recotnmend that expedmenters open them-
subtly and probably unconsciously influence tbe cbildten's perfor- selves to otbet potential benefits of pretend play. Well-being and
mance. Tbe biasing does not always bappen, perhaps because of a sense of petsonal agency ate two possibilities. Wben cbildren ate
stdcter expedmenters, procedures tbat are less vulnerable to bias, pretending tbey appear to feel in control. Tbis tnigbt be mitigated
or expedmenters witb less of tbe play etbos, but it happens enough in many expedments, wben cbildten are instructed to pretend, but
to taint tbe picture. Masked expedmenters are tare in cognitive expedmental conditions tnigbt be designed tbat minimize tbe sense
development researcb, but in tbis domain, not using tbem has been of extemal control. Anotbet issue to considet is tbe reverse of wbat
sbown to be a problem. Tbus, a clear recommendation from tbis was investigated bete: wbetbet absence of play is harmful, as bas
review is tbat in ptetend play researcb, masked expedmenters been sbown in some animals (Pellis & Pellis, 2009).
sbould be used unless measures are absolutely impervious to bias.
Witb mote expedments incorporating dgorous metbodology,
Those wbo administet outcome measures sbould not know bow
one day researchers will have a mote solid answer regarding
mucb cbildren have pretended in observations ot wbetbet cbildten
wbetber pretending belps specific aspects of cbildren's develop-
were in tbe ptetend play condition. Ideally intetvention adminis-
ment, is just one route among many, or simply often goes along
trators are masked to hypotheses, altemate conditions, and wbat
witb otber circumstances tbat lead to positive developments.
tests will be given.
Samples and conditions. In expedmental teseatcb, random Conclusion
assignment to condition is necessary. If cbildten cboose tbeir
condition (acting classes) tben unmeasured preexisting differences Despite over 40 years of researcb examining bow pretend play
could account for results. Further, care must be taken to develop migbt belp development, tbere is littie evidence tbat it bas a crucial
control conditions tbat ate equal except for tbe pretend element: role; equifinality and epipbenomenalism have as mucb if not mote
Tbe degree of adult interaction, tbe content, and tbe context must support. Witb equifinality, ptetend play would be just one of many
be the same save tbe ptetend play. Finally, researcbers must avoid routes to a positive developmental outcome. Witb epipbenomenal-
potential confounds witb otber stable vadables tbat migbt influ- ism, ptetend play would often go along witb a positive develop-
ence cbildten's development, for example, by providing several mental outcome, but for extraneous reasons; it would not itself
implementets pet condition, ot using implementers as tbeir own serve any causal role in tbat outcome.
control but without tbem baving a personal stake in tbe perfor- Because the Utetattire is dddled witb weak metbods (correlational
mance of cbildren in one condition but not another. studies, lack of masked expedmenters, poor control conditions) atid
Measures. Reseatcbets should use unifotm measures to fa- utmgotous statistical apptoacbes, we cannot definitively state wbicb
cilitate compadsons across studies. For younger cbildren, McCu- of these models is most supported. In many areas tbe current reseatcb
ne's bierarcby (dedved from Piaget; see footnote 10) is a useful base is clearly inconsistent with tbe causal model, but leaves open tbe
standard. For examining natural play in tbe classroom, a service- othet two. Tbe methodological problems must be temedied witb
able and widely used scodng system nests Patten's (1932) in sound expedments and longitudinal studies before we can know
Smilansky's (1968) scbeme, as descdbed in tbe Creativity section wbetber and how ptetend play helps development. Meanwhile, tbe
(Rubin, 2001). We would also add to tbis combined scbeme a tally lack of existing evidence tbat ptetend play belps development sbould
of the number of transformations cbildren make. If pretending not be taken as an allowance fot school programs to employ tradi-
assists children because it gets tbem tbinking in unusual ways, tben tional teachet-centeted instructional approaches tbat researcb bas
tracking tbe extent to wbicb cbildren do so is important. When not clearly sbown ate itifedot for young cbildren. Tbe bands-on, cbild-
coding naturally occutdng play, using a standard measure is im- ddven educational metbods sometimes referred to as "playful leam-
28 LILLARD ET AL.
ing" (Hirsh-Pasek et al., 2009) are the most positive means yet known Berlyne, D. (1960). Conflict, arousal, and curiosity. New York, NY:
to help young children's development. McGraw Hill.
Blair, C , & Raver, C. C. (2012). Individual development and evolution:
References Experiential canalization of self-regulation. Developmental Psychology,
48, 647-657. doi:10.1037/a0026472
Aisenson, N. (1978). Fantasy and conservation among second grade chil- Blair, C , & Razza, R. P. (2007). Relating effortful control, executive
dren. The Journal of Genetic Psychology, 132, 155-156. function, and false belief understanding to emerging math and literacy
Albertson, K., & Shore, C. (2009). Holding in mind conflicting infor- ability in kindergarten. Child Development, 78, 647-663. doi:10.1111/
mation: Pretending, working memory, and executive control. Journal J.1467-8624.2007.01019.X
of Cognition and Development, 9, 390-410. doi:10.1080/ Bodrova, E., & Leong, D. (1996). Tools of the mind: The Vygotskian
15248370802678240 approach to early childhood education. Englewood Cliffs, NJ: Merrill.
Alfieri, L., Brooks, P. J., Aldrich, N. J., & Tenenbaum, H. R. (2011). Does Bonawitz, E., Shafto, P., Gweon, H., Goodman, N. D., Spelke, E., &
discovery-based instruction enhance leaming? Journal of Educational Schulz, L. (2011). The double-edged sword of pedagogy: Instruction
Psychology, 103, 1-18. doi:10.1037/a0021017
limits spontaneous exploration and discovery. Cognition, 120, 322-330.
Altman, D. G., & Bland, J. M. (1995). Statistics notes: Absence of doi:10.1016/j.cognition.2010.10.001
evidence is not evidence of absence. British Medical Journal, 311, 485.
Borke, H. (1971). Interpersonal perception of young children: Egocentrism
doi:10.1136/bmj.311.7003.485
or empathy? Developmental Psychology, 5, 263-269. doi: 10.1037/
Ashiabi, G. S. (2007). Play in the preschool classroom: Its socioemotional h0031267
significance and the teacher's role in play. Early Childhood Education
Borke, H. (1975). Piaget's mountains revisited: Changes in the egocentric
Journal, 35, 199-207. doi:10.1007/sl0643-007-0165-8
landscape. Developmental Psychology, 11, 240-243. doi:10.1037/
Astington, J. W., Harris, P. L., & Olson, D. R. (Eds.). (1988). Developing
h0076459
theories of mind. New York, NY: Cambridge University Press.
Bomstein, M. H., Vibbert, M., Tal, J., & O'Donnell, K. (1992). Toddler
Astington, J. W., & Jenkins, J. M. (1995). Theory of mind development
language and play in the second year: Stability, covariation and influ-
and social understanding. Cognition & Emotion, 9, 151-165. doi:
ences of parenting. First Language, 12, 323-338. doi:10.1177/
10.1080/02699939508409006
014272379201203607
Austin, J. L. (1979). Pretending. In J. O. Urmson & G. J. Wamock (Eds.),
Bosacki, S., & Astington, J. W. (1999). Theory of mind in preadolescence:
Philosophical papers (3rd ed., pp. 253-271). Oxford, England: Oxford
Relations between social understanding and social competence. Social
University Press.
Development, 8, 237-255. doi:10.mi/1467-9507.00093
Banta, T. J. (1970). Tests for the evaluation of early childhood education:
Brainerd, C. J. (1982). Effects of group and individualized dramatic play
The Cincinnati Autonomy Test Battery (CATB). In J. Hellmuth (Ed.),
Cognitive studies (pp. 424-490). New York, NY: Bumner Mazel. training on cognitive development. In D. J. Pepler & K. H. Rubin (Eds.),
The play of children: Current theory and research (pp. 114-129). Basel,
Bamett, L. A. (1984). Research note: Young children's resolution of
distress through play. Journal of Child Psychology and Psychiatry, 25, Switzerland: Karger.
477-483. doi:10.111 l/j.l469-7610.1984.tb00165.x Bredekamp, S. (2004). Play and school readiness. In E. F. Zigler, D. G.
Bamett, L. A., & Storm, B. (1981). Play, pleasure, and pain: The reduction Singer, & S. J. Bishop-Josef (Eds.), Children's play: The roots of
of anxiety through play. Leisure Sciences, 4, 161-175. doi:10.1080/ reading (pp. 159-174). Washington, DC: Zero to Three/National Center
01490408109512958 for Infants, Toddlers and Famihes.
Bamett, W., Jung, K., Yarosz, D., Thomas, J., Hombeck, A., Stechuk, R., Bretherton, I. (1989). Pretense: The form and function of make-believe
& Bums, S. (2008). Educational effects of the Tools of the Mind play. Developmental Review, 9, 383-401. doi: 10.1016/0273-
curriculum: A randomized trial. Early Childhood Research Quarterly, 2297(89)90036-1
23, 299-313. doi:10.1016/j.ecresq.2008.03.001 Brown, S., & Vaughan, C. (2009). Play: How it shapes the brain, opens the
Bates, E. (1979). The emergence of symbols: Cognition and communica- imagination, and invigorates the soul. New York, NY: Penguin.
tion in infancy. New York, NY: Academic Press. Buchsbaum, D., Gopnik, A., Griffiths, T. L., & Shafto, P. (2011). Chil-
Bates, E., Bretherton, I., Snyder, L., Shore, C , & Volterra, V. (1980). dren's imitation of causal action sequences is influenced by statistical
Vocal and gestural symbols at 13 months. Merrill-Palmer Quarterly, 26, and pedagogical evidence. Cognition, 120, 331-340. doi:10.1016/
407-423. j.cognition.2010.12.001
Bateson, G. A. (1972). A theory of play and fantasy. In G. A. Bateson Burghardt, G. M. (2011). Defining and recognizing play. In A. Pellegrini
(Ed.), Steps to an ecology of mind (pp. 177-193). New York, NY: (Ed.), Oxford handbook of the development of play (pp. 9-18). New
Chandler. York, NY: Oxford University Press.
Baumer, S., Ferholt, B., & Lecusay, R. (2005). Promoting narrative com- Bums, S. M., & Brainerd, C. J. (1979). Effects of constructive and dramatic
petence through adult-child joint pretense: Lessons from the Scandina- play on perspective-taking in very young children. Developmental Psy-
vian educational practice of playworld. Cognitive Development, 20, chology, 15, 512-521. doi:10.1037/0012-1649.15.5.512
576-590. doi:10.1016/j.cogdev.2005.08.003 Carlson, S., White, R. E., & Davis-Unger, A. (2012). Evidence for a
Benson, M. S. (1993). The structure of four- and five-year-olds' narratives relationship between executive function and pretense representation in
in pretend play and storytelling. First Language, 13, 203-223. doi: preschool children. Manuscript submitted for publication.
10.1177/014272379301303803 Casby, M. W., & Corte, M. D. (1987). Symbolic play performance and
Bergen, D. (2002). The role of pretend play in children's cognitive devel- early language development. Journal of Psycholinguistic Research, 16,
opment. Early Childhood: Research and Practice, 4, 1-13. 31-42.
Bergen, D. (2009). Play as the leaming medium for future scientists, Cemore, J. J., & Herwig, J. E. (2005). Delay of gratification and make-
mathematicians, and engineers. American Journal of Play, 1, 413-428. believe play of preschoolers. Journal of Research in Early Childhood
Berk, L. W., Mann, T. D., & Ogan, A. T. (2006). Make-believe play: Education, 19, 251-267. doi:10.1080/02568540509595069
Wellspring for development of self-regulation. In D. G. Singer, R. M. Chandler, M. (1973). Egocentrism and antisocial behavior: The assessment
Golinkoff, & K. Hirsh-Pasek (Eds.), Play = learning: How play moti- and training of social perspective-taking skills. Developmental Psychol-
vates and enhances children's cognitive and social-emotional growth ogy, 9, 326-332. doi:10.1037/h0034974
(pp. 74-100). New York, NY: Oxford University Press. Chandler, M., Greenspan, S., & Barenboim, C. (1974). Assessment and
PLAY DEVELOPMENT 29
training of role-taking and referential communication skills in institu- de Lorimier, S., Doyle, A., & Tessier, O. (1995). Social coordination
tionalized emotionally disturbed children. Developmental Psychology, during pretend play: Comparisons with nonpretend play and effects on
10, 546-553. doi:10.1037/h0036735 expressive contexts. Merrill-Palmer Quarterly, 41, 497-516.
Cheyne, J. A. (1982). Object play and problem-solving: Methodological Diamond, A., Bamett, W., Thomas, J., & Munro, S. (2007, November 30).
problems and conceptual promise. In D. J. Pepler & K. H. Rubin (Eds.), Preschool program improves cognitive control. Science, 318, 1387-
The play of children: Current theory and research (pp. 79-96). Basel, 1388. doi: 10.1126/science.l 151148
Switzerland: Karger. Diamond, A., & Lee, K. (2011, August 19). How can we help children
Cheyne, J. A., & Rubin, K. H. (1983). Playful precursors of problem succeed in the 21st century? What the scientific evidence shows aids
solving in preschoolers. Developmental Psychology, 19, 577-584. doi: executive function development in children 4-12 years of age. Science,
10.1037/0012-1649.19.4.577 333, 959-964. doi: 10.1126/scienee. 1204529
Chien, N. C , Howes, C , Burchinal, M., Pianta, R. C , Ritchie, S., Bryant, Dias, M. G., & Harris, P. L. (1988). The effect of make-believe play on
D. M Barbarin, O. A. (2010). Children's classroom engagement and deductive reasoning. British Journal of Developmental Psychology, 6,
school readiness gains in prekindergarten. Child Development, 81, 207-221. doi: 10.111 l/j.2044-835X.1988.tb01095.x
1534-1549. doi:10.1111/j.l467-8624.2010.01490.x Dias, M. G., & Harris, P. L. (1990). The infiuence of the imagination on
Christakis, D. A., Zimmerman, F. J., & Garrison, M. M. (2007). Effect of reasoning by young children. British Journal of Developmental Psychol-
block play on language acquisition and attention in toddlers: A pilot ogy, 8, 305-318. doi:10.1111/j.2044-835X.1990.tb00847.x
randomized controlled trial. Archives of Pediatrics & Adolescent Med- Doekett, S. (1998). Constructing understandings through play in the early
icine, 161, 967-971. doi:10.1001/archpedi.l61.10.967 years. International Journal of Early Years Education, 6, 105-116.
Christie, J. F. (1983). The effects of play tutoring on young children's doi:10.1080/0966976980060109
cognitive performance. The Journal of Educational Research, 76, 326- Doswell, G., Lewis, V., Sylva, K., & Boucher, J. (1994). Validational data
330. on the Warwick Symbolic Play Test. European Journal of Disorders of
Christie, J. F., & Johnsen, E. P. (1983). The role of play in social- Communication, 29, 289-298. doi:10.3109/13682829409111613
intellectual development. Review of Educational Research, 53{\), 9 3 - Doyle, A., Ceschin, F., Tessier, O., & Doehring, P. (1991). The relation of
115. doi: 10.3102/00346543053001093 age and social class factors in children's social pretend play to cognitive
Clements, D. H., Sarama, J., Unlu, F., & Layzer, C. (2012, March). The and symbolic activity. International Journal of Behavioral Develop-
efficacy of an intervention synthesizing scaffolding designed to promote ment, 14, 395-410.
self-regulation with an early mathematics curriculum: Effects on exec- Doyle, A., & Connolly, J. (1989). Negotiation and enactment in social
utive function. Paper presented at the Society for Research on Educa- pretend play: Relations to social acceptance and social cognition. Early
tional Effectiveness, Washington, DC. Childhood Research Quarterly, 4, 289-302. doi: 10.1016/0885-
Cole, D., & LaVoie, J. C. (1985). Fantasy play and related cognitive 2006(89)90015-X
development in 2- to 6-year-olds. Developmental Psychology, 21, 233- Doyle, A., Doehring, P., Tessier, O., de Lorimier, S., & Shapiro, S. (1992).
240. doi:10.1037/0012-1649.21.2.233 Transitions in children's play: A sequential analysis of states preceding
Colwell, M. J., & Lindsey, E. W. (2005). Preschool children's pretend and and following social pretense. Developmental Psychology, 28, 137-144.
physical play and sex of play partner: Connections to peer competence. doi:10.1037/0012-1649.28.1.137
Sex Roles, 52, 497-509. doi: 10.1007/sl 1199-005-3716-8 Dunn, J., & Cutting, A. L. (1999). Understanding others, and individual
Connolly, J. A., & Doyle, A.-B. (1984). Relation of social fantasy play to differences in friendship interactions in young children. Social Devel-
social competence in preschoolers. Developmental Psychology, 20, 797- opment, 8, 201-219. doi:10.1111/1467-9507.00091
806. doi:10.1037/0012-1649.20.5.797 Dunn, J., & Hughes, C. (2001). "I got some swords and you're dead!":
Connolly, J. A., Doyle, A. B., & Reznik, E. (1988). Social pretend play and Violent fantasy, antisocial behavior, friendship, and moral sensibility in
social interaction in preschoolers. Journal of Applied Developmental young children. Child Development, 72, 491-505.
Psychology, 9, 301-313. doi: 10.1016/0193-3973(88)90032-9 Dunn, L., & Herwig, J. E. (1992). Play behaviors and convergent and
Copple, C , & Bredekamp, S. (2009). Developmentally appropriate prac- divergent thinking skills of young children attending full-day preschool.
tice in early childhood programs serving children from birth through Child Study Journal, 22, 23-38.
age 8. Washington, DC: National Association for the Education of Elias, C. L., & Berk, L. E. (2002). Self-regulation in young children: Is
Young Children. there a role for sociodramatic play? Early Childhood Research Quar-
Corrigan, R. (1982). The control of animate and inanimate components in terly, 17, 216-238. doi:10.1016/S0885-2006(02)00146-l
pretend play and language. Child Development, 53, 1343-1353. doi: Elkind, D. (2007). The power of play. Cambridge, MA: Da Capo Lifelong
10.2307/1129025 Books.
Dansky, J. L. (1980a). Cognitive consequences of sociodramatic play and Erikson, E. H. (1950). Childhood and society. New York, NY: Norton.
exploration training for economically disadvantaged preschoolers. Jour- Ervin-Tripp, S. E. (1991). Play in language development. In B. E. Scales,
nal of Child Psychology and Psychiatry, 21, 47-58. doi:10.111 l/j.1469- M. E. Almy, A. E. Nicolopoulou, & S. E. Ervin-Tripp (Eds.), Play and
7610.198O.tbO0O15.x the social context of development in early care and education (pp.
Dansky, J. L. (1980b). Make-believe: A mediator of the relationship 84-97). New York, NY: Teachers College Press.
between play and associative fluency. Child Development, 51, 576-579. Fantuzzo, J., Sekino, Y., & Cohen, H. (2004). An examination of the
doi: 10.2307/1129296 contributions of interactive peer play to salient classroom competencies
Dansky, J. L., & Silverman, I. W. (1973). Effects of play on associative for urban Head Start children. Psychology in the Schools, 41, 323-336.
fluency In preschool-aged children. Developmental Psychology, 9, 3 8 - doi:10.1002/pits.l0162
43. doi:10.1037/h0035076 Fein, G. G. (1981). Pretend play in childhood: An integrative review. Child
Dansky, J. L., & Silverman, I. W. (1975). Play: A general facilitator of Development, 52, 1095-1118. doi:10.2307/l 129497
associative fluency. Developmental Psychology, II, 104. doi:10.1037/ Fein, G. G. (1989). Mind, meaning, and affect: Proposals for a theory of
hOO761O8 pretense. Developmental Review, 9, 345-363. doi:10.1016/0273-
DeLoache, J. S. (2000). Dual representation and young children's use of 2297(89)90034-8
scale models. Child Development, 71, 329-338. doi:10.1111/1467- Fein, G. G., & Kinney, P. (1994). He's a nice alligator: Observations on the
8624.00148 affective organization of pretense. In A. Slade & D. P. Wolf (Eds.),
30 LE.LARD ET AL.
Children at play: Clinical and developmental approaches to meaning Golomb, C , & Adams, J. (1978). Pretense play: A study of its cognitive
and representation (pp. 188-205). New York, NY: Oxford University significance. Unpublished manuscnpt. University of Massachusetts,
Press. Boston, MA.
Feitelson, D., & Ross, G. (1973). The neglected factor: Play. Human Golomb, C , & Bonen, S. (1981). Playing games of make-believe: The
Development, 16, 202-223. doi: 10.1159/000271276 effectiveness of symbolic play training with children who failed to
Fink, R. S. (1976). Role of imaginative play in cognitive development. benefit from early conservation training. Genetic Psychology Mono-
Psychological Reports, 39, 895-906. doi:10.2466/piO.I976.39.3.895 graphs, 104, 137-159.
Fisher, E. P. (1992). The impact of play on development: A meta-analysis. Golomb, C , & Comelius, C. B. (1977). Symbolic play and its cognitive
Play & Culture, 5, 159-181. significance. Developmental Psychology, 13, 246-252. doi: 10.1037/
Fisher, K. R., Hirsh-Pasek, K., Golinkoff, R. M., & Gryfe, S. G. (2008). 0012-1649.13.3.246
Conceptual split? Parents' and experts' perceptions of play in the 21st Golomb, C , Gowing, E. D. G., & Friedman, L. (1982). Play and cognition:
century. Journal of Applied Developmental Psychology, 29, 305-316. Studies of pretence play and conservation of quantity. Journal of Ex-
doi:10.1016/j.appdev.2008.04.006 perimental Child Psychology, 33, 257-279. doi: 10.1016/0022-
Flavell, J. H. (1978). The development of knowledge about visual percep- 0965(82)90019-4
tion. In C. B. Keasey (Ed.), Nebraska Symposium on Motivation: Human Gredlein, J. M., & Bjorklund, D. F. (2005). Sex differences in young
emotion (Vol. 26, pp. 43—76). Lincoln, NE: University of Nebraska children's use of tools in a problem-solving task. Human Nature, 16,
Press. 211-232. doi:10.1007/sl2110-005-1004-5
Flavell, J. H. (1988). The development of children's knowledge about the Guthrie, K., & Hudson, L. H. (1979). Training conservation through
mind: From cognitive connections to mental representations. In J. W. symbolic play: A second look. Child Development, 50, 1269-1271.
Astington, P. L. Harris, & D. R. Olson (Eds.), Developing theories of doi:10.2307/1129366
mind (pp. 244-267). Cambridge, England: Cambridge University Press. Harris, P. L. (1995). From simulation to folk psychology: The case for
Flavell, J., Botkin, P., Fry, C , Wright, J., & Jarvis, P. (1968). The development. In M. Davies & T. Stone (Eds.), Folk psychology (Vol. 3,
development of role-taking and communication skills in children. New pp. 207-221). Cambridge, England: Blackwell.
York, NY: Wiley. Harris, P. L. (2000). The work of the imagination. Oxford, England:
Flavell, J. H., Flavell, E. R., & Green, F. L. (1983). Development of the Blackwell.
appearance-reality distinction. Cognitive Psychology, 15, 95-120. doi: Harris, P. L., & Leevers, H. J. (2000). Reasoning from false premises. In
10.1016/0010-0285(83)90005-1 P. Mitchell & K. Riggs (Eds.), Children's reasoning and the mind (pp.
Forguson, L., & Gopnik, A. (1988). The ontogeny of common sense. In 67-86). East Sussex, England: Psychology Press.
J. W. Astington, P. L. Harris, & D. R. Olson (Eds.), Developing theories Hawkins, J., Pea, R. D., Glick, J., & Scribner, S. (1984). "Merds that don't
of mind (pp. 226-243). New York, NY: Cambridge University Press. like mushrooms": Evidence for deductive reasoning by preschoolers.
Fredrickson, B. L. (2001). The role of positive emotions in positive Developmental Psychology, 20, 584-594. doi:10.1037/0012-
psychology: The broaden-and-build theory of positive emotions. Amer- 1649.20.4.584
ican Psychologist, 56, 218-226. doi:10.1037/0003-066X.56.3.218 Henniger, M. L. (1991). Play revisited: A critical element of the kinder-
Freud, S. (1955). Analysis of a phobia in a five-year-old boy. In Collected garten curriculum. Early Child Development and Care, 70, 63-71.
papers (Vol. 3). London, England: Hogarth. doi:10.1080/0300443910700107
Galyer, K., & Evans, 1. (2001). Pretend play and the development of Hirsh-Pasek, K., Golinkoff, R., Berk, L., & Singer, D. (2009). A mandate
emotion regulation in preschool children. Early Child Development and for playful learning in preschool: Presenting the evidence. New York,
Care, 166, 93-108. doi: 10.1080/0300443011660108 NY: Oxford University Press.
Garon, N., Bryson, S., & Smith, I. (2008). Executive function in preschool- Howard-Jones, P., Taylor, J., & Sutton, L (2002). The effect of play on the
ers: A review using an integrative framework. Psychological Bulletin, creativity of young children during subsequent activity. Early Child
134, 31-60. doi:10.1037/0033-2909.134.1.31 Development and Care, 172, 323-328. doi:10.1080/03004430212722
Garvey, C. (1990). Play (2nd ed.). Cambridge, MA: Harvard University Howes, C , & Matheson, C. C. (1992). Sequences in the development of
Press. competent play with peers: Social and social pretend play. Development
Garvey, C , & Kramer, T. L. (1989). The language of social pretend play. Psychology, 28, 961-974. doi:10.1037/0012-1649.28.5.961
Developmental Review, 9, 364-382. doi: 10.1016/0273-2297(89) Hughes, C , & Dunn, J. (1997). "Pretend you didn't know": Preschoolers'
90035-X talk about mental states in pretend play. Cognitive Development, 12,
Gaskins, S., & Goncu, A. (1992). Cultural variation in play: A challenge to 411-491. doi:10.1016/S0885-2014(97)90019-8
Piaget and Vygotsky. The Quarterly Newsletter of the Laboratory of Hurwitz, S. C. (2002). To be successful—Let them play! Childhood
Comparative Human Cognition, 14, 31-35. Education, 79, 101-102.
Giffin, H. (1984). The coordination of meaning in the creation of shared Hutt, S. J., Tyler, S., Hutt, C , & Christcpherson, H. (1989). Play, explo-
make-believe play. In 1. Bretherton (Ed.), Symbolic play (pp. 73-100). ration, and leaming: A natural history of the preschool. London,
Orlando, FL: Academic Press. England: Routledge.
Ginsburg, K. R., the Committee on Communications, & the Committee on Iannotti, R. J. (1978). Effect of role-taking experiences on role taking,
Psychosocial Aspects of Child and Family Health. (2007). The impor- empathy, altruism, and aggression. Developmental Psychology, 14,
tance of play in promoting healthy child development and maintaining 119-124. doi:10.1037/0012-1649.14.2.119
strong parent-child bonds. Pediatrics, 119, 182-191. doi:10.1542/ Ilgaz, H., & Aksu-Koç, A. (2005). Episodic development in preschool
peds.2006-2697 children's play-prompted and direct-elicited narratives. Cognitive De-
Goldstein, T. R., & Winner, E. (2010). Engagement in role play, pretense, velopment, 20, 526-544. doi:10.1016/j.cogdev.2005.08.004
and acting classes predict advanced theory of mind skill in middle Isenberg, J., & Quisenberry, N. L. (1988). Play: A necessity for all
childhood. Imagination, Cognition and Personality, 30, 249-258. doi: children. Childhood Education, 64, 138-145.
10.2190AC.30.3.C Jenkins, J. M., & Astington, J. W. (2000). Theory of mind and social
Goldstein, T. R., & Winner, E. (2012). Enhancing empathy and theory of behavior: Causal models tested in a longitudinal study. Merrill-Palmer
mind. Journal of Cognition and Development, 13, 19-37. doi:10.1080/ Quarterly, 46, 203-220.
15248372.2011.573514 Johnson, J. (1976). Relations of divergent thinking and intelligence test
PLAY DEVELOPMENT 31
scores with social and nonsoeial make-believe play of pteschool chil- Lillard, A. S. (2005). Montessori: The science behind the genius. New
dren. Child Development, 47, 1200-1203. doi: 10.2307/1128465 York, NY: Oxford University Press.
Johnson, J. (1978). Mother-child interaction and imaginative behavior of Lillard, A. (2011). Mother-child fantasy play. In A. D. Pelligrini (Ed.), The
preschool children. Journal of Psychology, ¡00, 123-129. doi:10.1080/ Oxford handbook of the development of play (pp. 284-295). New York,
00223980.1978.9923480 NY: Oxford University Press.
Johnson, J., Etshler, J., & Lawton, J. (1982). Intellective correlates of Lillard, A. S. (2012). Preschool children's development in classic Montes-
preschoolers' spontaneous play. Journal of General Psychology, 106, sori, supplemented Montessori, and conventional programs. Journal of
115-122. doi:10.1080/00221309.1982.9710980 School Psychology, 50, 379-401. doi:10.1016/j.jsp.2012.01.001
Jurkovic, G. (1978). Relation of psycholinguistic development to imagi- Lillatd, A., & Else-Quest, N. (2006, September 29). Evaluating Montessori
native play of disadvantaged preschool children. Psychology in the education. Science, 313, 1893-1894. doi:10.1126/science.l 132362
Schools, 15, 560-564. doi:10.1002/1520-6807(197810)15.4<560::AID- Lillard, A. S., & Erisir, A. (2011). Old dogs learning new tricks: Neuro-
PITS2310150420>3.0.CO;2-3 plasticity before and after critical periods. Developmental Review, 31,
Kavanaugh, R. D. (2006). Pretend play and theoty of mind. In L. Baiter & 207-239. doi:10.1016/j.dr.2011.07.008
C. S. Tamis-LeMonda (Eds.), Child psychology: A handbook of contem- Lillatd, A. S., & Peterson, J. (2011). The immediate impact of different
porary issues (pp. 153-166). Philadelphia, PA: Psychology Press. types of television on young children's executive function. Pediatrics,
Kavanaugh, R. D., & Lillard, A. S. (2012). Predicting theory of mind from 128, 644-649. doi:10.1542/peds.2010-1919
symbolic understanding, language, and pretend play. Williams College, Lillard, A. S., Pinkham, A., & Smith, E. D. (2011). Pretend play and
Williamstown, MA. cognitive development. In U. Goswami (Ed.), Handbook of cognitive
Kelly, R., Hammond, S., Dissanayake, C , & Ihsen, E. (2011). The tela- development (2nd ed., pp. 285-311). London, England: Blackwell.
tionship between symbolic play and executive function in young chil- Lindsey, E. W., & Colwell, M. J. (2003). Preschoolers' emotional compe-
dren. Australasian Journal of Early Childhood, 36, 21-28. tence: Links to ptetend and physical play. Child Study Journal, 33,
Kim, S. Y. (1999). The effects of storytelling and pretend play on cognitive 39-52.
processes, shott-term and long-term narrative recall. Child Study Jour- Lloyd, B., & Howe, N. (2003). Solitary play and convergent and divergent
nal, 29, 175-191. thinking skills in pteschool children. Early Childhood Research Quar-
Klahr, D., & Nigam, M. (2004). The equivalence of leaming paths in early terly, 18, 22-41. doi:10.1016/S0885-2006(03)00004-8
science instruction: Effects of direct instruction and discovery learning. Lonigan, C. J., & Phillips, B. M. (2012, March). Comparing skills-focused
Psychological Science, 15, 661-667. doi:10.1111/J.0956- and self-regulation focused preschool curricula: Impacts on academic
7976.2004.00737.x and self-regulatory skills. Paper presented at the Society for Research on
Krasnor, L. R., & Pepler, D. J. (1980). The study of children's play: Some Educational Effectiveness, Washington, DC.
suggested future directions. In K. H. Rubin (Ed.), Children's play: New Lovinger, S. L. (1974). Socio-dramatic play and language development
directions for child development (pp. 85-95). San Francisco, CA: in preschool disadvantaged children. Psychology in the Schools, 11,
Jossey-Bass. 313-320. doi:10.1002/1520-6807(197407)ll:3<313::AID-
Kuezaj, S. A. (1981). Factors influencing children's hypothetical ref- PITS2310110315>3.0.CO,2-L
erence. Journal of Child Language, 8, 131-137. doi:10.1017/ Lyytinen, P., Laakso, M. L., Poikkeus, A. M., & Rita, N. (1999). The
S0305000900003056 development and predictive relations of play and language actoss the
Lalonde, C , & Chandler, M. (1995). False belief understanding goes to second year. Scandinavian Journal of Psychology, 40, 177-186. doi:
school: On the social-emotional consequences of coming early or late to 10.1111/1467-9450.00115
a first theory of mind. Cognition & Emotion, 9, 167-185. doi:10.1080/ Lyytinen, P., Poikkeus, A.-M., & Laakso, M.-L. (1997). Language and
02699939508409007 symbolic play in toddlers. International Journal of Behavioral Devel-
Lancy, D. F. (2007). Accounting for variability in mother-child play. opment, 21, 289-302. doi:10.1080/016502597384875
American Anthropologist, 109, 273-284. doi:10.1525/aa.2007.109.2.273 Manuilenko, Z. (1975). The development of voluntary behavior in
Leslie, A. M. (1987). Pretense and representation: The origins of "theoty of preschool-age children. Journal of Russian and East European Psychol-
mind." Psychological Review, 94, 412-426. doi: 10.1037/0033- ogy, 13, 65-116. doi:10.2753/RPO1061-0405130465
295X.94.4.412 Marbach, E. S., & Yawkey, T. D. (1980). The effect of imaginative play
Levy, A. K., Schaefet, L., & Phelps, P. C. (1986). Increasing pteschool actions on language development in five-year-old children. Psychology
effectiveness: Enhancing the language abilities of 3- and 4-year-old in the Schools, 17, 257-263. doi:10.1002/1520-6807(198004)17:
children through planned sociodtamatic play. Early Childhood Research 2<257::AID-PITS2310170218>3.0.CO;2-A
Quarterly, 1, 133-140. doi:10.1016/0885-2006(86)90024-4 Matthews, W. S., Beebe, S., & Bopp, M. (1980). Spatial petspecüve-taking
Lewis, V., & Boucher, J. (1997). The test of pretend play. London, and pretend play. Perceptual and Motor Skills, 51, 49-50. doi: 10.2466/
England: Psychological Corporation. pms. 1980.51.1.49
Lewis, V., Boucher, J., Lupton, L., & Watson, S. (2000). Relationships Mayer, R. E. (2004). Should there be a three-strikes rule against pure
between symbolic play, functional play, verbal and non-verbal ability in discovery leaming? American Psychologist, 59, 14-19. doi:10.1037/
young children. International Journal of Language & Communication 0003-066X.59.1.14
Disorders, 35, 117-127. doi:10.1080/136828200247287 MeAloney, K., & Stagnitti, K. (2009). Pretend play and social play: The
Li, A. K. (1978). Effects of play on novel responses in kindergarten concurrent validity of the Child-Initiated Pretend Play Assessment.
children. Alberta Journal of Educational Research, 24, 31-36. International Journal of Play Therapy, 18, 99-113. doi: 10.1037/
Lillard, A. S. (1993). Ptetend play skills and the child's theory of mind. a0014559
Child Development, 64, 348-371. doi:10.2307/113I255 McCune, L. (1995). A normative study of representational play in the
Lillard, A. S. (2001a). Pretend play as Twin Earth: A social-cognitive transition to language. Developmental Psychology, 31, 198-206. doi:
analysis. Developmental Review, 21, 495-531. doi: 10.1006/ 10.1037/0012-1649.31.2.198
drev.2001.0532 McLoyd, V. (1983). The effects of the structure of play objects on the
Lillard, A. S. (2001b). Pretending, understanding pretense, and understand- ptetend play of low-income preschool children. Child Development, 54,
ing minds. In S. Reifel (Ed.), Play and culture studies (Vol. 3, pp. 626-635. doi:10.2307/l 130049
233-254). Westport, CT: Ablex. Méndez, J. L., & Fogle, L. M. (2002). Parental reports of preschool
32 LILLARD ET AL.
children's social behavior: Relations among peer play, language com- play paradigm. American Educational Research Journal, 21, 691-701.
petence, and problem behavior. Journal of Psychoeducational Assess- doi: 10.2307/1162926
ment, 20, 370-385. doi:10.1177/073428290202000405 Pellegrini, A. (2009). The role of play in human development. New York,
Mikami, A. Y. (2010). The importance of friendship for youth with NY: Oxford University Press. doi:10.1093/acprof:oso/9780195367324
attention-deficit/hyperactivity disorder. Clinical Child and Family Psy- .001.0001
chology Review, 13, 181-198. doi:10.1007/sl0567-010-0067-y Pellegrini, A., & Bohn, C. (2005). The role of recess in children's cognitive
Miller, E., & Almon, J. (2009). Crisis in the kindergarten: Why children performance and school adjustment. Educational Researcher, 34, 13-19.
need to play in school. College Park, MD: Alliance for Childhood. doi:10.3102/0013189X034001013
Mischel, W., Shoda, Y., & Rodriguez, M. L. (1989, May 26). Delay of Pellegrini, A., & Galda, L. (1982). The effects of thematic-fantasy play
gratification in children. Science, 244, 933-938. doi: 10.1126/ training on the development of children's story comprehension. Ameri-
science.2658056 can Educational Research Journal, 19, 443-452. doi: 10.2307/1162724
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, Pellegrini, A., & Greene, H. (1980). The use of a sequenced questioning
A. (2000). The unity and diversity of executive functions and their paradigm to facilitate associative fluency in preschoolers. Journal of
contributions to complex "frontal lobe" tasks: A latent variable analysis. Applied Developmental Psychology, I, 189-200. doi: 10.1016/0193-
Cognitive Psychology, 41, 49-100. doi:10.1006/cogp.l999.0734 3973(80)90008-8
Montessori, M. (1989). The child, society, and the world: Unpublished Pellegrini, A., & Gustafson, K. (2005). Boys' and girls' uses of objects for
speeches and writings (Vol. 7). Oxford, England: Clio. exploration, play, and tools in early childhood. In A. Pellegrini & P.
Moore, M., & Russ, S. (2008). Follow-up of a pretend play intervention: Smith (Eds.), The nature of play: Great apes and humans (pp. 113-135).
Effects on play, creativity, and emotional processes in children. Cre- New York, NY: Guilford Press.
ativity Research Journal, 20, 427-436. doi:10.1080/ Pellegrini, A., & Smith, P. (1993). School recess: Implications for educa-
10400410802391892 tion and development. Review of Educational Research, 63, 51-67.
Moran, J., Sawyers, J., Fu, V., & Milgram, R. (1984). Predicting imagi- Pellis, S., & Pellis, V. (2009). The playful brain: Venturing to the limits of
native play in preschool children. Gifted Child Quarterly, 25, 92-94. neuroscience. Oxford, England: Oneworld.
doi:10.1177/001698628402800209
Pepler, D. J., & Ross, H. S. (1981). The effects of play on convergent and
Moses, L., & Chandler, M. (1992). Traveler's guide to children's theories
divergent problem solving. Child Development, 52, 1202-1210. doi:
of mind. Psychological Inquiry, 3, 286-301. doi:10.1207/
10.2307/1129507
sl5327965pliO3O3_22
Phillips, R. D. (2010). How firm is our foundation? Current play therapy
Neisser, U., Boodoo, G., Bouchard, T. J., Jr., Boykin, A. W., Brody, N.,
research. International Journal of Play Therapy, 19, 13-25. doi:
Ceci, S. J Urbina, S. (1996). Intelligence: Knowns and unknowns.
10.1037/a0017340
American Psychologist, 51, 77-101. doi:10.1037/0003-066X.51.2.77
Piaget, J. (1929). The child's conception of the world. London, England:
Neuman, S. B., & Roskos, K. (1992). Literacy objects as cultural tools:
Routledge & Kegan Paul.
Effects on children's literacy behaviors in play. Reading Research
Piaget, J. (1932). The moral judgement of the child. New York, NY:
Quarterly, 27, 203-225. doi: 10.2307/747792
Harcourt, Brace.
Nielsen, M., & Dissanayake, C. (2000). An investigation of pretend play,
Piaget, J. (1962). Play, dreams, and imitation in childhood (G. Gattegno &
mental state terms and false belief understanding: In search of a metarep-
F. M. Hodgson, Trans.). New York, NY: Norton.
resentational link. British Journal of Developmental Psychology, 18,
609-624. doi:10.1348/026151000165887 Power, T. G. (2000). Play and exploration in children and animals.
Nisbett, R. E. (2009). Intelligence and how to get it: Why schools and Mahwah, NJ: Erlbaum.
cultures count. New York, NY: Norton. Premack, D., & Woodnifl", G. (1978). Does the chimpanzee have a theory
Ogura, T. (1991). A longitudinal study of the relationship between early of mind? Behavioral and Brain Sciences, 1, 515-526. doi:10.1017/
language development and play development. Journal of Child Lan- S0140525X00076512
guage, 18, 273-294. doi:10.1017/S0305000900011065 Putnam, H. (1996). The meaning of "meaning." In A. Pessin & S. Goldberg
O'Neill, D. K., & Shultis, R. (2007). The emergence of the ability to track (Eds.), The Twin Earth chronicles (pp. 3—52). London, England: Sharpe.
a character's mental perspective in narrative. Developmental Psychol- (Original work published 1975)
ogy, 43, 1032-1037. doi:10.1037/0012-1649.43.4.1032 Ramani, G. B., & Siegler, R. S. (2008). Promoting broad and stable
Overton, W. F., & Jackson, J. P. (1973). The representation of imagined improvements in low-income children's numerical knowledge through
objects in action sequences: A developmental study. Child Development, playing number board games. Child Development, 79, 375-394. doi:
44, 309-314. doi: 10.2307/1128052 10.1 lll/j.1467-8624.2007.01131.x
Paley, V. G. (1990). The boy who would be a helicopter. Cambridge, MA: Ratey, J. (2008). Spark: The revolutionary new theory of exercise. New
Harvard University Press. York, NY: Little, Brown.
Paley, V. G. (2005). A child's work: The importance of fantasy play. Richards, C. A., & Sanderson, J. A. (1999). The role of imagination in
Chicago, YL\ University of Chicago Press. facilitating deductive reasoning in 2-, 3- and 4-year-olds. Cognition,
Parten, M. (1932). Social participation among preschool children. The 72(2), B1-B9. doi:10.1016/S0010-0277(99)00037-2
Journal of Abnormal and Social Psychology, 27, 243-269. doi: 10.1037/ Roopnarine, J. L. (2011). Cultural variations in beliefs about play, parent-
hO074524 child play, and children's play: Meaning for childhood development. In
Parten, M. (1933). Social play among preschool children. The Journal of A. Pellegrini (Ed.), Oxford handbook of the development of play (pp.
Abnormal and Social Psychology, 28, 136-147. doi:10.1037/h0073939 19-37). New York, NY: Oxford University Press.
Peisach, E., & Hardeman, M. (1985). Imaginative play and logical thinking Root-Bernstein, M. M. (in press). The creation of imaginary worlds. In M.
in young children. The Journal of Genetic Psychology, 146, 233-248. Taylor (Ed.), The oxford handbook of the imagination. New York, NY:
doi:10.1080/00221325.1985.9914451 Oxford University Press.
Pellegrini, A. (1981). A sequenced questioning paradigm as a general Rosen, C. (1974). The effects of sociodramatic play on problem-solving
facilitator of preschoolers' associative fluency. Perceptual and Motor behavior among culturally disadvantaged preschool children. Child De-
Skills, 52, 649-650. doi: 10.2466/pms. 1981.52.2.649 velopment, 45, 920-927. doi:10.1111/1467-8624.epl2116929
Pellegrini, A. (1984). Identifying causal elements in the thematic-fantasy Roskos, K., & Neuman, S. (1998). Play as an opportunity for literacy. In
PLAY DEVELOPMENT 33
O. N. Saracho & B. Spodek (Eds.), Multiple perspectives on play in story familiarity, and adult intervention. Merrill-Palmer Quarterly, 32,
early childhood education (pp. 100-115). Albany, NY: SUNY Press. 73-86.
Rubin, K. H. (1980). Fantasy play: Its role in the development of social Simon, T., & Smith, P. K. (1983). The study of play and problem solving
skills and social cognition. In K. H. Rubin (Ed.), Children's play: New in preschool children: Have experimenter effects been responsible for
directions for child development (Vol. 9, pp. 69-84). San Francisco, previous results? British Journal of Developmental Psychology, I, 289-
CA: Jossey-Bass. 297. doi:10.1111/J.2044-835X.1983.tb00901.x
Rubin, K. H. (1982). Nonsocial play in preschoolers: Necessarily evil? Simon, T., & Smith, P. K. (1985). Play and problem solving: A paradigm
Child Development, 53, 651-657. doi: 10.2307/1129376 questioned. Merrill-Palmer Quarterly, 31, 265-277.
Rubin, K. H. (2001). The Play Observation Scale (POS). Unpublished Singer, D. G., Golinkoff, R. M., & Hirsh-Pasek, K. (Eds.). (2006). Play =
manuscript. University of Maryland. leaming: How play motivates and enhances children's cognitive and
Rubin, K. H., & Daniels-Beimess, T. (1983). Concurrent and predictive social-emotional growth. New York, NY: Oxford University Press.
correlates of sociometric status in kindergarten and Grade 1 children. Singer, D. G., & Singer, J. L. (1992). The house of make believe: Chil-
Merrill-Palmer Quarterly, 29, 337-351. dren's play and the developing imagination. Cambridge, MA: Harvard
Rubin, K. H., & Maioni, T. L. (1975). Play preference and its relationship University Press.
to egocentrism, popularity, and classification skills in preschoolers. Singer, D. G., Singer, J. L., D'Agostino, H., & Delong, R. (2009). Chil-
Merrill-Palmer Quarterly, 21, 171-179. dren's pastimes and play in sixteen nations: Is free play declining?
Rubin, K. H., & Pepler, D. J. (1980). The relationship of child's play to American Journal of Play, I, 283-312.
social-cognitive growth and development. In H. C. Foot, A. J. Chapman, Singer, J. L. (1961). Imagination and waiting ability in young children.
& J. R. Smith (Eds.), Friendship and social relations in children (pp. Journal of Personality, 29, 3 9 6 - 4 1 3 . doi:10.111 l/j.1467-
209-233). New York, NY: Wiley. 6494.1961.tb01670.x
Rueda, M. R., Posner, M. I., & Rothbart, M. K. (2005). The development Singer, J. L. (1973). The child's world of make-believe. New York, NY:
of executive attention: Contributions to the emergence of self-regulation. Academic Press.
Developmental Neuropsychology, 28, 573-594. doi:10.1207/ Smilansky, S. (1968). The effects of sociodramatic play on disadvantaged
preschool children. New York, NY: Wiley.
sl5326942dn2802_2
Smith, E. D., & Lillard, A. S. (in press). Play on: Retrospective evidence
Russ, S. W., & Grossman-McKee, A. (1990). Affective expression in
for the persistence of pretend play into middle childhood. Journal of
children's fantasy play, primary process thinking on the Rorschach, and
Cognition and Development.
divergent thinking. Journal of Personality Assessment, 54, 756-771.
Smith, P. K. (1988). Children's play and its role in early development: A
doi:10.1207/sl5327752jpa5403&4_28
re-evaluation of the "play ethos." In A. D. Pellegrini (Ed.), Psycholog-
Russ, S. W., & Kaugars, A. S. (2001). Emotion in children's play and
ical bases for early education (pp. 207-226). New York, NY: Wiley.
creative problem solving. Creativity Research Journal, 13, 211-219.
Smith, P. K. (2010). Children and play. West Sussex, England: Wiley-
doi: 10.1207/S15326934CRJ1302_8
Blackwell.
Russ, S. W., Robins, A. L., & Christiano, B. A. (1999). Pretend play:
Smith, P. K., Dalgleish, M., & Herzmark, G. (1981). A comparison ofthe
Longitudinal prediction of creativity and affect in fantasy in children.
effects of fantasy play tutoring and skills tutoring in nursery classes.
Creativity Research Journal, 12, 129-139. doi:10.1207/ International Journal of Behavioral Development, 4, 421-441.
sl5326934crjl202_5
Smith, P. K., & Dutton, S. (1979). Play and training in direct and innova-
Saltz, E., Dixon, D., & Johnson, J. (1977). Training disadvantaged pre-
tive problem solving. Child Development, 50, 830-836. doi:10.2307/
schoolers on various fantasy activities: Effects on cognitive function and
1128950
impulse control. Child Development, 48, 367-380. doi: 10.2307/1128629
Smith, P. K., Simon, T., & Emberton, R. (1985). Play, problem solving and
Saltz, E., & Johnson, J. (1974). Training for thematic-fantasy play in experimenter effects: A replication of Simon & Smith (1983). British
culturally disadvantaged children: Preliminary results. Journal of Edu- Journal of Developmental Psychology, 3, 105-107. doi:10.111 l/j.2044-
cational Psychology, 66, 623-630. doi:10.1037/h0036930 835X.1985.tb00961.x
Sattler, J. (1992). The assessment of children: WISC-III and WPPSI-R Smith, P. K., & Syddall, S. (1978). Play and non-play tutoring in preschool
supplement. San Diego, CA: Sattler. children: Is it play or tutoring that matters? British Journal of Educa-
Schellenberg, E. G. (2004). Music lessons enhance IQ. Psychological tional Psychology, 48, 315-325. doi:10.11 n/j.2044-8279.1978
Science, 15, 511-514. doi:10.1111/j.0956-7976.2004.00711.x .tbO3O17.x
Schwebel, D., Rosen, C , & Singer, J. (1999). Preschoolers' pretend play Smith, P. K., & Whitney, S. (1987). Play and associative fluency: Exper-
and theory of mind: The role of jointly constructed pretence. British imenter effects may be responsible for previous positive findings. De-
Journal of Developmental Psychology, 17, 333-348. doi: 10.1348/ velopmental Psychology, 23, 49-53. doi:10.1037/0012-1649.23.1.49
026151099165320 Stagnitti, K., & Unsworth, C. (2000). The importance of pretend play in
Scott, C. L., Harris, R. J., & Rothe, A. R. (2001). Embodied cognition child development: An occupational therapy perspective. British Journal
through improvisation improves memory for a dramatic monologue. of Occupational Therapy, 63, 121-127.
Discourse Processes, 31, 293-305. doi:10.1207/S15326950dp31-3_4 Stipek, D., Feiler, R., Daniels, D., & Milbum, S. (1995). Effects of
Shields, A., & Cicchetti, D. (1997). Emotion regulation among school-age different instructional approaches on young children's achievement and
children: The development and validation of a new criterion Q-sort motivation. Child Development, 66, 209-223. doi:10.2307/l 131201
scale. Developmental Psychology, 33, 906-916. doi:10.1037/0012- Suddendorf, T., Fletcher-Flinn, C , & Johnston, L. (1999). Pantomime and
1649.33.6.906 theory of mind. The Journal of Genetic Psychology, 160, 31-45. doi:
Shore, C. (1986). Combinatorial play, conceptual development, and early 10.1080/00221329909595378
multiword speech. Developmental Psychology, 22, 184-190. doi: Sutton-Smith, B. (1966). Piaget on play: A critique. Psychological Review,
10.1037/0012-1649.22.2.184 73, 104-110. doi:10.1037/h0022601
Shore, C , O'Connell, B., & Bates, E. (1984). First sentences in language Sutton-Smith, B. (1967). The role of play in cognitive development. Young
and symbolic play. Developmental Psychology, 20, 872-880. doi: Children, 22, 361-370.
10.1037/0012-1649.20.5.872 Sutton-Smith, B. (1968). Novel responses to toys. Merrill-Palmer Quar-
Silvem, S. B. (1986). Young children's story recall as a product of play. terly, 14, 151-158.
34 LILLARD ET AL.
Sutton-Smith, B. (1995). Conclusion: The persuasive rhetorics of play. In Vandenberg, B. (1981). The role of play in the development of insightful
A. D. Pellegrini (Ed.), The future of play theory: A multidisciplinary tool-using strategies. Merrill-Palmer Quarterly, 27, 97-109.
inquiry into the contributions of Brian Sutton-Smith (pp. 275-295). Vandenberg, B. (1990). Play and problem-solving: An elusive connection.
Albany, NY: SUNY Press. Merrill-Palmer Quarterly, 36, 261-272.
Sutton-Smith, B., & Kelly-Byrne, D. (1984). The idealization of play. In Vygotsky, L. S. (1967). Play and its role in the mental development of the
P. K. Smith (Ed.), Play in animals and humans (pp. 305-321). Oxford, child. Soviet Psychology, 5, 6-18.
England: Blackwell. Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard Uni-
Swindells, D., & Stagnitti, K. (2006). Pretend play and parents' view of versity Press.
social competence: The construct validity of the Child-Initiated Pretend Walker, R. F., & Murachver, T. (2012). Representation and theory of mind
Play Assessment. Australian Occupational Therapy Journal, 53, 314- development. Developmental Psychology, 48, 509-520. doi:10.1037/
324. doi:10.1111/j.l440-1630.2006.00592.x a0025663
Sylva, K. (1974). The relation between play and problem-solving in Wallach, M. A., & Kogan, N. (1965). A new look at the creativity-
children 3 to 5 years of age (Doctoral thesis). Harvard University, intelligence distinction. Journal of Personality, 33, 348-369. doi:
Cambridge, MA. 10.1111/j.l467-6494.1965.tb01391.x
Sylva, K. (1977). Play and learning. In B. Tizard & D. Harvey (Eds.), Watson, A. C , Nixon, C. L., Wilson, A., & Capage, L. (1999). Social
Biology of play (pp. 59-73). London, England: SIMP/Heinemann. interaction skills and theory of mind in young children. Developmental
Sylva, K., Bruner, J., & Genova, P. (1976). The role of play in the Psychology, 35, 386-391. doi:10.1037/0012-1649.35.2.386
problem-solving of children 3-5 years old. In A. J. J. Bruner & K. Sylva Wellman, M. (1990). The child's theory of mind. Cambridge, MA: Brad-
(Ed.), Play: Its role in development and evolution (pp. 244-257). New ford Books/MIT Press.
York, NY: Basic Books. Wenner, M. (2009, January 28). The serious need for play. Scientific American
Tamis-LeMonda, C , & Bomstein, M. (1994). Specificity in mother- Mind, 20, 22-29. doi:10.1038/scientificamericanniind0209-22
toddler language-play relations across the second year. Developmental Wemer, H., & Kaplan, B. (1963). Symbol formation. New York, NY:
Psychology, 30, 283-292. doi:10.1037/0012-1649.30.2.283 Wiley.
Taylor, M., & Carlson, S. M. (1997). The relation between individual Whitebread, D., Bingham, S., Grau, V., Pino Pasternak, D., & Sangster, C.
differences in fantasy and theory of mind. Child Development, 68, (2007). Development of metacognition and self-regulated learning in
436-455. doi:10.2307/1131670 young children: Role of collaborative and peer-assisted learning. Journal
Taylor, M., Carlson, S. M., Maring, B. L., Gerow, L., & Charley, C. M. of Cognitive Education and Psychology, 6, 433-455. doi: 10.1891/
(2004). The characteristics and correlates of fantasy in school-age chil- 194589507787382043
dren: Imaginary companions, impersonation, and social understanding. Williamson, P. A., & Silvern, S. B. (1990). The effects of play training on
Developmental Psychology, 40, 1173-1187. doi:10.1037/0012- the story comprehension of upper primary children. Journal of Research
1649.40.6.1173 in Childhood Education, 4, 130-134. doi:10.1080/02568549009594794
Tenenbaum, H. R., Snow, C. E., Roach, K. A., & Kurland, B. (2005). Wilson, R. C , Guilford, J., & Christensen, P. R. (1953). The measurement
Talking and reading science: Longitudinal data on sex differences in of individual differences in originality. Psychological Bulletin, 50, 362-
mother-child conversations in low-income families. Journal of Applied 370. doi:10.1037/h0060857
Developmental Psychology, 26, 1-19. doi: 10.1016/j.appdev.2004 Wilson, S. J., & Farran, D. C. (2012, March). Experimental evaluation of
.10.004 the Tools of the Mind preschool curriculum. Paper presented at the
Tomlinson, H. B. (2009). Developmentally appropriate practice in the Society for Research on Educational Effectiveness, Washington, DC.
kindergarten year—Ages 5-6. In C. Copple & S. Bredecamp (Eds.), Wimmer, H., & Pemer, J. (1983). Beliefs about beliefs: Representation and
Developmentally appropriate practice in early childhood programs (pp. constraining function of wrong beliefs in young children's understand-
186-216). Washington, DC: National Association for the Education of ing of deception. Cognition, 13, 103-128. doi:10.1016/0010-
Young Children. 0277(83)90004-5
Torrance, E. P. ( 1966). Thinking creatively with pictures: Eigurat booklet. Wyver, S. R., & Spence, S. H. (1999). Play and divergent problem solving:
Bensenville, IL: Scholastic Testing Service. Evidence supporting a reciprocal relationship. Early Education and
Torrance, E. P. (1981). Thinking creatively in action and movement. Development, 10, 419-444. doi:10.1207/sl5566935eedl004_l
Bensenville, IL: Scholastic Testing Service. Youngblade, L. M., & Dunn, J. (1995). Individual differences in young
Trionfi, G., & Reese, E. (2009). A good story: Children with imaginary children's pretend play with mother and sibling: Links to relationships
companions create richer narratives. Child Development, 80, 1301- and understanding of other people's feelings and beliefs. Child Devel-
1313. doi:10.1111/j.l467-8624.2009.01333.x opment, 66, 1472-1492. doi:10.2307/l 131658
Tullis, P. (2011, November 2). Preschool tests take time away from Ziegler, F., Mitchell, P., & Curde, G. (2005). How does narrative cue
play—and learning. Scientific American Mind. Retrieved from http:// children's perspective taking? Developmental Psychology, 41, 115—123.
www.scientificamerican.com/article.cfm?id=the-death-of-presehool doi:10.1037/0012-1649.41.1.115
Ungerer, J. A., & Sigman, M. (1984). The relation of play and sensorimotor Zigler, E., & Bishop-Josef, S. (2004). Play under siege: A histodcal
behavior to language in the second year. Child Development, 55, 1448- overview. In E. F. Zigler, D. G. Singer, & S. J. Bishop-Josef (Eds.),
1455. doi:10.2307/1130014 Children's play: The roots of reading (pp. 1-13). Washington, DC: Zero
Uren, N., & Stagnitti, K. (2009). Pretend play, social competence and to Three/National Center for Infants, Toddlers and Families.
involvement in children aged 5-7 years: The concurrent validity of the
Child-Initiated Pretend Play Assessment. Australian Occupational Ther-
apy Journal, 56, 33-40. doi:10.1111/j.l440-1630.2008.00761.x Received December 9, 2011
Vandenberg, B. (1980). Play, problem solving, and creativity. New Direc- Revision received June 4, 2012
tions for Child and Adolescent Development, 9, 49-68. Accepted June 6, 2012
Copyright of Psychological Bulletin is the property of American Psychological Association and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.