Language and Conceptual Development series
Psychological essentialism in children
Susan A. Gelman
University of Michigan, Ann Arbor, MI 48109-1109, USA
Psychological essentialism is the idea that certain
categories, such as ‘lion’ or ‘female’, have an underlying
reality that cannot be observed directly. Where does this
idea come from? This article reviews recent evidence
suggesting that psychological essentialism is an early
cognitive bias. Young children look beyond the obvious
in many converging ways: when learning words,
generalizing knowledge to new category members,
reasoning about the insides of things, contemplating
the role of nature versus nurture, and constructing
causal explanations. These findings argue against the
standard view of children as concrete thinkers, instead
claiming that children have an early tendency to search
for hidden, non-obvious features.
Why do preschool children often insist that mothers can’t
be firefighters? Why do adults who were adopted in
infancy sometimes search for their birth parents? Why
do art collectors pay more money for an original painting
than for an exact copy? These examples, although widely
varying in content, can be understood within a framework
of psychological essentialism.
Essentialism is the view that certain categories have an
underlying reality or true nature that one cannot observe
directly but that gives an object its identity, and is
responsible for other similarities that category members
share [1,2]. In the domain of biology, an essence would be
whatever quality remains unchanging as an organism
grows, reproduces, and undergoes morphological transformations
(baby to man; caterpillar to butterfly). In the
domain of chemistry, an essence would be whatever
quality remains unchanging as a substance changes
shape, size, or state (from solid to liquid to gas).
Essentialist accounts have been proposed and discussed
for thousands of years, extending back at least to
Plato’s cave allegory in The Republic. Numerous fields,
including biology, philosophy, linguistics, literary criticism,
and psychology, stake claims about essentialism.
Here we are concerned with essentialism as a psychological
claim (see Box 1). Although there are serious
problems with essentialism as a metaphysical doctrine,
recent psychological studies converge to suggest that
essentialism is a reasoning heuristic that is readily
available to both children and adults [3]. This article
reviews such evidence, and discusses the implications for
human concepts.
Evidence for psychological essentialism
Fodor argued that essentialism is the outgrowth of
modern scientific and technological knowledge, with its
corresponding focus on non-obvious constructs such as
genes and chromosomes [4]. Children provide a critical
test case of this idea precisely because they lack detailed
scientific knowledge. If essentialism requires knowledge
of modern science and technology then it should emerge
relatively late in development. By contrast, if essentialism
can be found in preschool children then this would imply
that the doctrine is likely to have a more fundamental
basis.
What would be evidence for essentialism, in children or
adults? Medin and Ortony suggest that essentialism is a
‘placeholder’ notion: one can believe that a category
possesses an essence without knowing what the essence
is [2]. For example, a child might believe that there exist
deep, non-visible differences between males and females,
but have no idea just what those differences are. The
essence placeholder would imply that categories permit
rich inductive inferences, capture underlying structure (in
the form of causal and other non-obvious properties), have
innate potential, and have sharp and immutable boundaries.
I have detailed at length elsewhere the evidence
that preschool children expect certain categories to have
all of these properties [3]. It is beyond the scope of this
article to review the full range of evidence here. However,
I briefly summarize below some of the major findings.
Inductive potential
Induction is the capacity to extend knowledge to novel
instances, for example, inferring that a newly encountered
mushroom is poisonous on the basis of past encounters
with other poisonous mushrooms. This capacity is one of
the most important functions of categories [5]. Categories
serve not only to organize the knowledge we have already
acquired but also to guide our expectations. Young
children’s category-based inferences are consistent with
essentialism in two respects. First, children readily infer
properties that concern internal features and non-visible
functions from one category member to another. Second,
children draw inferences even when category membership
competes with perceptual similarity.
Figure 1 provides an example from a set of studies
conducted some years ago by Gelman and Markman [6,7].
The leaf (Figure 1a) and the leaf-insect (Figure 1c) have
overall similarity: both are large and green, with striped
markings, and share overall shape. However, if told the
category membership of each of the three items – (a) ‘leaf’,
Corresponding author: Susan A. Gelman (gelman@umich.edu).
Available online 29 July 2004
www.sciencedirect.com 1364-6613/$ - see front matter Q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.tics.2004.07.001
Review TRENDS in Cognitive Sciences Vol.8 No.9 September 2004
(b) ‘bug’, (c) ‘bug’ – and asked to draw novel inferences
about the leaf-insect, children rely on the category
membership conveyed by the label. Once children learn
a new fact about one member of a category, they generalize
the fact to other members of that category, even if the two
category members look substantially different. This effect
holds up for animals (bird, fish, rabbit), for natural
substances (gold, cotton), and for social categories (boy,
girl, smart, shy) [8,9]. It is found when using threedimensional
objects as well as drawings [10]. The effect is
found with adults as well as children [6,11]. Even 1- and
2-year-old children draw category-based inferences about
atypical instances [12,13]. Thus, the appreciation that
words can signal non-obvious properties seems in place at
the very start of word learning.
By 4 years of age, children display subtlety and
flexibility regarding when they do and don’t make
category-based inductive inferences [3]. They do not use
a simple matching strategy, in which they extend properties
only when two items share identical labels. The effect
emerges not only with familiar labels, although not
uniformly with all labels. The explanation that accounts
most satisfactorily for the varied patterns of data is that
children assess the extent to which entities are members
of the same category (often conveyed via a label or phrase,
although not necessarily), and independently assess the
extent to which the property in question is a relatively
enduring (versus temporary or accidental) feature (but see
[14] for debate). Category-based induction results only
when the entities belong to a category and the property is
relatively enduring. When other causal mechanisms are
known and available, however, category-based induction
seems to be less frequent [5].
Innate potential
One of the most important kinds of evidence for essentialism
is the belief that properties are fixed at birth, that
is, that an organism displays innate potential. Details
vary, but the basic paradigm to test this notion is as
follows: children learn about a person or animal that has
a set of biological parents, but is switched at birth to a
new environment and a new set of parents. Children are
then asked to decide whether the birth parents or the
upbringing parents determine various properties.
For example, in one item set, children learned about a
newborn kangaroo that went to live with goats, and were
asked whether it would be good at hopping or good at
climbing, and whether it would have a pouch or no pouch
[15]. Preschool children typically reported that it would be
good at hopping and have a pouch. Even if it cannot hop at
birth (because it is too small and weak), and is raised by
goats that cannot hop, and never sees another kangaroo,
hopping is inherent to kangaroos. Therefore this property
Box 1. Defining essentialism
The words ‘essence’ and ‘essentialism’ are used loosely by a range of
scholars to mean a wide range of things. Three factors jointly serve to
map out the various types of essentialism [43].
Where is the essence located?
Is it in the world (metaphysical essentialism) or in human representations
(representational essentialism)? The former is a philosophical
position concerning objective reality; the latter addresses how people
construe reality (in their belief systems, language, and cultural
practices). Essentialism faces difficulties as a characterization of the
natural world [2]. Social categories, such as race and caste, have no
true underlying essence [17,21]. Biological species also are without an
essence because they evolve and are population-based rather than
reflecting properties inherent in each individual [44,45]. Furthermore,
whereas essentialism implies that there is a single appropriate
classification for each organism, there might in fact be numerous
valid classifications [46]. The essentialist view therefore seems to be a
human construction rather than a perceived reality.
What is the ontological type of an essence?
Is it sortal (serving to define categories), causal (having consequences
for category structure), or ideal (having no real-world instantiation)?
The sortal essence is the set of defining characteristics that all and
only members of a category share. On this view the essence of a
grandmother would be the property of being the mother of a person’s
parent (rather than accidental properties such as having gray hair).
The viability of this account has been called into question by models of
concepts that stress the importance of probabilistic features, exemplars,
and theories in concepts [38].
The causal essence is the substance, power, quality, process,
relationship, or entity that causes other category-typical properties
to emerge and be sustained, and that confers identity. For example,
the causal essence of tigers might be something like tiger DNA, which
is responsible for various observable properties that tigers have. The
cluster of properties ‘striped, ferocious, and large’ is not a causal
essence of tigers, despite being true of all (or most) members of the
category, because the properties lack causal force.
The ideal essence is assumed to have no actual instantiation in the
world. For example, on this view the essence of ‘goodness’ is some
pure quality that is imperfectly realized in real-world instances of
people performing good deeds. Plato’s cave allegory, in which what
we see of the world are mere shadows of what is real and true,
exemplifies this view. The ideal essence contrasts with both the sortal
and the causal essences, which concern qualities of real-world
entities.
What degree of specificity is entailed?
Are essences specific (their particulars known and identified) or
placeholder (their particulars unknown and perhaps unknowable)?
Specific essentialism posits that a category essence is known and
contributes to the meaning and use of a category label (e.g. H2O for
water), whereas placeholder essentialism suggests that a person
believes that there is some causal essence that holds a category
together, without knowing just what that essence is (e.g. that all
samples of water share some inherent, non-obvious property).
Although placeholder essentialism would typically be insufficient to
determine word extensions, it has implications for people’s beliefs
regarding the depth and stability of a concept.
In this article, the focus is on representational, causal, placeholder
essentialism.
(a) (b) (c)
Figure 1. Sample item set used in studies with preschool children [6,7]. The item
sets were constructed so that overall similarity (in shape and color) was pitted
against shared category membership. This set comprised (a) a leaf, (b) a beetle, and
(c) a leaf-insect. When 3- and 4-year-old children heard labels for these items (‘leaf’,
‘bug’, and ‘bug’, respectively), they were more likely to extend new information on
the basis of shared label than on the basis of overall similarity.
Review TRENDS in Cognitive Sciences Vol.8 No.9 September 2004 405
www.sciencedirect.com
will eventually be expressed. Although there is debate as
to when precisely this understanding emerges [16], on a
conservative estimate it appears by about 6 years of age,
and in some studies as early as age 4 years. This is so when
children reason about animal categories, plant categories,
and social categories [3,17,18]. Intriguingly, for some
categories children are more nativist than adults. For
example, 5-year-olds predict that a child who is switched
at birth will speak the language of the birth parents rather
than the adoptive parents [19]. The degree to which a
nativist bias is expressed varies across cultures, but
Torguud adults in Western Mongolia [20], adults in
India [21], Vezo children in Madagascar [22], and Itzaj
Maya adults and children in Mexico [23] all display a
nativist bias.
Underlying structure
Underlying structure can be seen in the importance
children grant to causality, and to non-obvious features.
By 2 years of age children view causes as vital to what
something is. Features that are causes are more core than
non-causal features [24]. Causes are more important than
mere associations; causes are also more important than
effects [25]. Furthermore, features that are causally
coherent (meaning that they fit together in an explanatory
framework) are weighted more heavily than features
that are equally available but not participating in a
network of causal explanations. Recent work with adults
details more formally the importance of causality in
adults’ concepts [26,27].
Causality is further central to children’s categories in
that children provide consistent, domain-specific causal
explanations for the properties that members of a category
share. Even 3-year-olds attribute an animal’s actions to
the animal itself, rather than to an external force
(inherent cause) [28]. By 4 years of age, children appeal
to ‘energy’ as a causal force [3,29]. By 8 years of age,
children attribute actions and growth to an animal’s
‘insides’.
Children also privilege internal, non-obvious properties
in their categories (see Figures 2 and 3). By 4 or 5 years of
age, children often recognize that an animal cannot be
transformed into another kind of thing (for example, a
raccoon cannot become a skunk). Instead, category
membership is stable over striking transformations [15,30].
Also, non-obvious properties, especially internal properties,
appear to be salient to young children and are
privileged in their determinations of what things are [15].
This finding appears not only among middle-class children
in the US, but also in middle-class and Favela-dwelling
children in Brazil [31].
Summary
The studies reviewed here suggest that both preschool
children and adults from a variety of cultural contexts
expect members of a category to be alike in non-obvious
ways. They treat certain categories as having inductive
potential, an innate basis, stable category membership,
and sharp boundaries between contrasting categories.
These beliefs are not the result of a detailed knowledge
base, nor are they imparted directly by parents [32],
although language might play an important indirect role
(see Box 2). Instead, they appear early in childhood with
relatively little direct prompting.
What psychological essentialism implies about children
Childhood essentialism poses a challenge to more traditional
theories of children’s concepts, which have emphasized
their focus on superficial, accidental, or perceptual
Figure 2. Sample item used to study what procedures children deem relevant when
determining ambiguous category membership (Lizotte and Gelman, reported in
[3]). Five-year-olds and college students saw a series of trials depicting a pair of
items that looked nearly identical. They were told that the two items differed in
some respect (e.g. one was a dog and one was a wolf; or one was an animal and one
was a toy), and that their job was to figure out which item was which. They were
asked to decide (yes/no answers) whether checking any of the following would be
informative procedures: insides, origins/parentage, behaviors, or age. The results
indicated that children and adults believe that the items can be distinguished by a
wide range of means: not just by external behaviors but also by internal properties
and origins. Five-year-olds and adults reported that origins and insides are
important clues as to which of two seemingly identical animals is a dog or a
wolf, although it is unlikely that they could say specifically how the insides of dogs
and wolves differ.
Figure 3. Sample items used in a study examining the role of internal parts in
directing children’s word learning [31]. (The original items were in color.)
(a) Brazilian and US 4-year-old children learned a new word referring to a novel
sub-type of a familiar category; for example, ‘zava’ for the snake in the bottom panel
of (a), and were then asked to find the snake(s) and the zava(s) from among the
three pictures in (b). Children who heard that the two animals in (a) shared internal
similarities (e.g. have the same kind of bones, blood, muscles and brain) were more
likely to learn the new word correctly (as a sub-type of snake) than children in a
control condition who heard that the two animals shared superficial similarities
(e.g. are the same size, and live in the same zoo in the same kind of cage). Children
in the control condition were instead more likely to interpret the two labels
(e.g. snake and zava) as mutually exclusive. Reproduced with permission from [31].
Review TRENDS in Cognitive Sciences Vol.8 No.9 September 2004406
www.sciencedirect.com
features. At the extreme, Piaget and his colleagues
suggested that children are incapable of forming ‘true’
concepts [33]. Likewise, many scholars have proposed one or
another developmental dichotomy: from concrete to
abstract, from surface to deep, from perceptual to conceptual.
On the contrary, there are remarkable commonalities
between the concepts of children and those of adults. Some
scholars still maintain that developmental dichotomies
might exist earlier in development (e.g. from 1–2 years,
instead of from 5–7 years [34], but see [35]).
However, essentialism does not posit that perceptual
features or similarity are unimportant to early concepts.
Evenwithinanessentialistframework,appearancesprovide
crucial cues regarding an underlying essence. Similarity
appears to play an important role in fostering comparisons
between representations and hence discovery of new
abstractions and regularities [36]. Rather than suggesting
that human concepts overlook perception or similarity,
essentialism carries with it the assumption that a category
has two distinct, although interrelated, levels: the level of
observable reality and the level of explanation and cause.
It is this two-tier structure that might serve to motivate
further development, leading children to develop more
sophisticated understandings. Most developmental
accounts of cognitive change include something like this
structure, such as equilibration, competition, theory
change, analogy, or cognitive variability (see [3] for
review). In all these cases, as with essentialism, children
consider contrasting representations. When new evidence
conflicts with the child’s current understanding, this can
lead the child gradually to construct new representations.
Indeed, targeted interventions that introduce a nonobvious
similarity between dissimilar things can lead to
dramatic change in children’s concepts [37]. Perhaps not
surprisingly, then, children look beyond observable features
when trying to understand the categories of their
world. In positing a reality beyond appearances, the
search is on for more information, deeper causes, and
alternative construals.
What psychological essentialism implies about
concepts
There is an idealized model of categorization that has
formed the basis for much work in psychology. Standard
theories of concepts have been based on considering which
known properties are most privileged, and in what form.
By contrast, essentialism tells us that known properties do
not constitute the full meaning of concepts. Concepts are
also open-ended. They are in part placeholders for
unknown properties.
Historically, it has often been assumed that there is a
single, unitary process of categorization [38]. Yet essentialism
makes clear that categorization cannot be said to
be a single thing. Categorization serves many different
functions, and we recruit different sorts of information
depending on the task at hand. Rapid identification calls
for one kind of process; reasoning about genealogy calls for
another. Task differences yield different categorization
processes [39]. Even when the task is restricted to object
identification, people make use of different sorts of
information depending on the task instructions [40]. Two
separate categorization procedures, rule application and
judging similarity to an exemplar, can readily apply to the
same categories, although they activate different neural
regions in the adult brain [41]. Domain is also an
important consideration. Although essence-like construals
can be found in concepts of artwork and artifacts
[42], essentialism is found most often with natural kinds
and certain social categories. Categories such as penguin
or apple tree imply inherent non-obvious properties,
inductive potential, and innate commonalities not found
with categories like window or crayon.
Because essentialism is found early in childhood, all of
the points discussed above must be operating in quite
basic and fundamental ways. In other words, we cannot
simply assume that these are frills added on top of basic or
standard categorization.
Conclusion
Recent work has shown that preschool children treat a
variety of natural categories as having substantial
inductive potential, innate properties, and underlying
structure. Related to this, there are several important
aspects to children’s concepts from very early in life,
including placeholder notions, theory-based properties,
category and task variability, and interdependence of
categorization and other cognitive processes. More generally,
these findings overturn assumptions about what is
Box 2. Language and essentialism
The language that is used to express membership in a category
can influence children’s judgments about that category. Count
nouns imply that a category is relatively more stable and
consistent over time and contexts than adjectives or verbal
phrases. For example, in one study [47], 5- and 7-year-old children
first learned about a set of individuals with either a count noun
(‘Rose is 8 years old. Rose eats a lot of carrots. She is a carroteater.’)
or a verbal phrase (‘Rose is 8 years old. Rose eats a lot of
carrots. She eats carrots whenever she can.’). They were then
probed for how stable they thought this category membership
would be across time and different environmental conditions
(e.g. ‘Will Rose eat a lot of carrots when she is grown up?’ ‘Would
Rose stop eating a lot of carrots if her family tried to stop her
from eating carrots?’). Children who heard the count noun were
more likely than children who heard the verbal phrase to judge
that the personal characteristics would be stable over time and
adverse environmental conditions. (For other examples of nounlabeling
effects, see also [12,13,48–50].)
Generic noun phrases express essential qualities and imply that a
category is coherent and permits category-wide inferences [51]. When
4-year-old children hear a new fact in generic form (e.g. ‘Bears have
three layers of fur’), they treat this fact as typically true of most or all
category members [52]. Generic nouns are plentiful in the speech that
children hear [32,53], and children are highly sensitive to cues that
mark genericity [54–56].
Additionally, there are language-specific devices that convey
essentialism. For example, young Spanish-speaking children make
inferences about the stability of a category based on which form of the
verb ‘to be’ is used to express it (ser versus estar; [57]). Although it is
unlikely that language is the source of psychological essentialism, it
provides important cues to children regarding when to treat categories
as stable and having an intrinsic basis [3].
Review TRENDS in Cognitive Sciences Vol.8 No.9 September 2004 407
www.sciencedirect.com
simple or basic in human concepts. I have provided a
framework of ‘psychological essentialism’ to account for
these data (although see Box 3 for alternative accounts
and debates). Ongoing research examines developmental
antecedents in infancy, how best to understand the
relation between perceptual and conceptual information
in children’s categories, and modeling these representations
in formal terms [24,26,27] (see Box 4).
Acknowledgements
Preparation of this article and some of the research described within were
supported by NICHD grant HD36043 to the author. I am grateful to Bruce
Mannheim, the Editor, and the anonymous reviewers for very helpful
comments on an earlier draft of the manuscript.
References
1 Locke, J. (1671/1959) An Essay Concerning Human Understanding
(Vol 2), Dover
2 Medin, D.L. and Ortony, A. (1989) Psychological essentialism. In
Similarity and Analogical Reasoning (Vosniadou, S. and Ortony, A.
eds), pp. 179–195, Cambridge University Press
3 Gelman, S.A. (2003) The Essential Child: Origins of Essentialism in
Everyday Thought, Oxford University Press
4 Fodor, J. (1998) Concepts: Where Cognitive Science Went Wrong,
Oxford University Press
5 Medin, D.L. et al. (2003) A relevance theory of induction. Psychon.
Bull. Rev. 10, 517–532
6 Gelman, S.A. and Markman, E.M. (1986) Categories and induction in
young children. Cognition 23, 183–209
7 Gelman, S.A. and Markman, E.M. (1987) Young children’s inductions
from natural kinds: the role of categories and appearances. Child Dev.
58, 1532–1541
8 Heyman, G.D. and Gelman, S.A. (2000) Preschool children’s use of
novel predicates to make inductive inferences about people. Cogn. Dev.
15, 263–280
9 Heyman, G.D. and Gelman, S.A. (2000) Preschool children’s use of traits
labels to make inductive inferences. J. Exp. Child Psychol. 77, 1–19
10 Dea´k, G. and Bauer, P.J. (1996) The dynamics of preschoolers’
categorization choices. Child Dev. 67, 740–767
Box 3. Alternative accounts
Although the data reviewed in this article are consistent with
psychological essentialism in children, they fall short of providing
direct evidence. Children do not explicitly articulate an essentialist
perspective, and the word ‘essence’ probably does not enter the
vocabulary until well into school-age years. Nor can we say with any
precision what the essence of, say, a dog or a tree is to a child.
Perhaps, then, essentialism is unnecessary, and we should instead
refer to the component phenomena discussed earlier (e.g. categories
have inductive potential; causal features are central to categories).
One reason to frame these results in terms of essentialism is
parsimony: a range of phenomena that co-occur within a relatively
brief developmental span seem to be instantiations of a single broader
principle. This argument, however, requires a closer empirical
foundation. More fine-grained analysis will be needed to discover
how tightly linked these various phenomena are, both in adult
cognition and in development (see [58] for an excellent start). A
second consideration is the existence of explicit essence formulations
among adults in varied cultures [3]. By treating people’s implicit
construals as essentialist, a common framework covers both the
explicit and the implicit phenomena.
Even if essentialist phenomena can reasonably be considered
interrelated, this does not mean that the concept is indivisible. Indeed,
psychological essentialism appears to have two related although
separable assumptions: (a) a kind assumption, that people treat
certain categories as richly structured ‘kinds’ with clusters of
correlated properties; and (b) an essence assumption, that people
believe a category has an underlying property (essence) that cannot be
observed directly but that causes the observable qualities that
category members share.
The major difference between kind and essence is that the latter
incorporates the former and adds to it the idea that some part or
quality (i.e. the essence) causes the properties shared by the kind. In
the literature, the two notions have often been treated as interchangeable.
However, categories can be bound together in crucial ways
without considering the causal basis of the kind. In an important
argument, Strevens [59] suggests that the data taken as evidence for
psychological essentialism could instead be accounted for if people
simply assume that there are causal laws connecting kind membership
with observable properties. He terms such causal laws ‘K-laws’
(kind laws), and his alternative formulation the ‘Minimal Hypothesis’.
Strevens’s account, although eschewing essentialism, overlaps with
the current model in emphasizing that people treat surface features as
caused and constrained by deeper features of concepts.
Developmentally, a notion of kind might precede a notion of essence.
Evidence for use of kinds is present by age 2 years (in children’s
inductive inferences), but evidence for appeal to an essence (e.g. with
the switched-at-birth method) has so far not appeared below age 4
years. An important direction for future research would be to use
converging methods to try to distinguish ‘kind’ from ‘essence’ at
different points in development. Nonetheless, something more than
the Minimal Hypothesis might be required to account for essentialistlike
behavior in children aged 4 years and over [60].
There are additional issues that are beyond the scope of this article
but nonetheless important to mention. Some scholars have argued
that essentialism cannot account for a variety of experimental findings
with adults [61–64]. For example, the extent to which different liquids
are judged to be water is independent of the extent to which they share
the purported essence of water, H2O. Whether these findings undermine
(or even conflict with) psychological essentialism is a matter of
current debate [3,39].
Box 4. Questions for future research
† How coherent are essentialist beliefs? Do different strands
(e.g. nativism, inductive potential, boundary intensification) all
‘hang together’, or do they develop piecemeal? Is there a single
essentialist ‘stance’ [65], or is it the amalgamation of a variety of other
tendencies [3]?
† Why do children (and adults) essentialize? Some existing
proposals include the following: an innate, domain-specific module
for reasoning about biological species [23]; a domain-general
consequence of sortals and count nouns [44]; the convergence of a
variety of other early-emerging capacities needed for conceptual
growth in childhood [3].
† We know that non-verbal and preverbal organisms categorize [34],
but do non-verbal and preverbal organisms also treat categories as
having a non-obvious basis? If so, how could we tell? If not, why not?
† How does essentialism change with the acquisition of more
detailed scientific knowledge? Does essentialism disappear with the
acquisition of a scientific theory, or do the two frameworks co-exist in
adult reasoning? Lay adults’ conceptual difficulties with evolutionary
theory and genetics argue against true conceptual reorganization [66].
† Can certain inputs reduce children’s reliance on essentialism? If
so, of what sort?
† Are there stable individual differences in essentialism, and if so,
what is the source of such differences [67,68]? Are individual
differences in essentialism related to individual differences in stereotyping?
How does essentialism differ from other related notions,
including stereotyping and entitativity [69]?
† Is essentialism found in the same way and to the same extent
across cultures, or is there systematic variation [70]?
Review TRENDS in Cognitive Sciences Vol.8 No.9 September 2004408
www.sciencedirect.com
11 Yamauchi, T. and Markman, A.B. (2000) Inference using categories.
J. Exp. Psychol. Learn. Mem. Cogn. 26, 776–795
12 Jaswal, V.K. and Markman, E.M. (2002) Children’s acceptance and
use of unexpected category labels to draw non-obvious inferences. In
Proceedings of the 24th Annual Conference of the Cognitive Science
Society (Gray, W. and Schunn, C. eds), pp. 500–505, Erlbaum
13 Graham, S.A. et al. (2004) Thirteen-month-olds rely on shared labels
and shape similarity for inductive inferences. Child Dev. 75, 409–427
14 Sloutsky, V.M. (2003) The role of similarity in the development of
categorization. Trends Cogn. Sci. 7, 246–251
15 Gelman, S.A. and Wellman, H.M. (1991) Insides and essences: early
understandings of the nonobvious. Cognition 38, 213–244
16 Solomon, G.E.A. (2002) Birth, kind, and naı¨ve biology. Dev. Sci. 5,
213–218
17 Hirschfeld, L. (in press) Children’s understanding of racial groups. In
Children’s Understanding of Society (Barrett, M. and BuchananBarrow,
E., eds), Psychology Press
18 Giles, J.W. (2003) Children’s essentialist beliefs about aggression. Dev.
Rev. 23, 413–443
19 Hirschfeld, L.A. and Gelman, S.A. (1997) What young children think
about the relation between language variation and social difference.
Cogn. Dev. 12, 213–238
20 Gil-White, F.J. (2001) Are ethnic groups biological ‘species’ to the
human brain? Curr. Anthropol. 42, 515–554
21 Mahalingam, R. (2003) Essentialism and beliefs about gender among
Aravanis, the ‘not-men’ of Tamil Nadu. Sex Roles 49, 489–499
22 Astuti, R. (2001) ‘Are we all natural dualists?: a cognitive developmental
approach. The Malinowski Memorial Lecture. J. R. Anthropol.
Inst. 7, 429–447
23 Atran, S. et al. (2001) Folkbiology doesn’t come from folkpsychology:
evidence from Yukatek Maya in cross-cultural perspective. J. Cogn.
Cult. 1, 3–42
24 Gopnik, A. et al. (2004) A theory of causal learning in children: causal
maps and Bayes nets. Psychol. Rev. 111, 3–32
25 Ahn, B. et al. (2000) Causal status effect in children’s categorization.
Cognition 76, B35–B43
26 Rehder, B. and Hastie, R. (2004) Category coherence and categorybased
proprety induction. Cognition 91, 113–153
27 Ahn, W. et al. (2000) Causal status as a determinant of feature
centrality. Cogn. Psychol. 41, 361–416
28 Gelman, R. (2002) Animates and other worldly things. In Representation,
Memory, and Development: Essays in Honor of Jean Mandler
(Stein, N. et al., eds), pp. 75–87, Erlbaum
29 Inagaki, K. and Hatano, G. (2002) Young Children’s Naı¨ve Thinking
About the Biological World, Psychology Press
30 Keil, F. (1989) Concepts, Kinds, and Cognitive Development, MIT
Press
31 Diesendruck, G. (2001) Essentialism in Brazilian children’s extensions
of animal names. Dev. Psychol. 37, 49–60
32 Gelman, S.A. et al. (1998). Beyond labeling: the role of parental input in
the acquisition of righly-structured categories. Monographs of the
Society for Research in Child Development Serial No. 253, Vol. 63, No. 1
33 Inhelder, B. and Piaget, J. (1964) The Early Growth of Logic in the
Child, Norton
34 Rakison, D.H. and Oakes, L.M. (2003) Early Category and Concept
Development: Making Sense of the Blooming, Buzzing Confusion,
Oxford University Press
35 Keil, F.C. et al. (1998) Two dogmas of conceptual empiricism:
implications for hybrid models of the structure of knowledge.
Cognition 65, 103–135
36 Namy, L.L. and Gentner, D. (2002) Making a silk purse out of two
sows’ ears: young children’s use of comparison in category learning.
J. Exp. Psychol. Gen. 131, 5–15
37 Opfer, J.E. and Siegler, R.S. Revisiting preschoolers’ living things
concept: a microgenetic analysis of conceptual change in basic biology.
Cogn. Psychol. (in press)
38 Murphy, G.L. (2002) The Big Book of Concepts, MIT Press
39 Rips, L.J. (2001) Necessity and natural categories. Psychol. Bull. 127,
827–852
40 Yamauchi, T. and Markman, A.B. (1998) Category learning by
inference and classification. J. Mem. Lang. 39, 124–148
41 Smith, E.E. et al. (1998) Alternative strategies of categorization.
Cognition 65, 167–196
42 Gelman, S.A. and Bloom, P. (2000) Young children are sensitive to how
an object was created when deciding what to name it. Cognition 76,
91–103
43 Gelman, S.A. and Hirschfeld, L.A. (1999) How biological is essentialism?
In Folkbiology (Medin, D.L. and Atran, S. eds), pp. 403–446, MIT Press
44 Mayr, E. (1991) One Long Argument: Charles Darwin and the Genesis
of Modern Evolutionary Thought, Harvard University Press
45 Sober, E. (1994) From a Biological Point of View, Cambridge
University Press
46 Dupre´, J. (1993) The Disorder of Things: Metaphysical Foundations of
the Disunity of Science, Harvard University Press
47 Gelman, S.A. and Heyman, G.D. (1999) Carrot-eaters and creaturebelievers:
the effects of lexicalization on children’s inferences about
social categories. Psychol. Sci. 10, 489–493
48 Walton, G.M. and Banaji, M.R. Being what you say: the effect of
essentialist linguistic labels on preferences. Social Cogn. (in press)
49 Waxman, S.R. (2003) Links between object categorization and
naming: origins and emergence in human infants. In Early Category
and Concept Development: Making Sense of the Blooming, Buzzing
Confusion (Rakison, D.H. and Oakes, L.M. eds), pp. 213–241, Oxford
University Press
50 Xu, F. (2002) The role of language in acquiring object kind concepts in
infancy. Cognition 85, 223–250
51 Prasada, S. (2000) Acquiring generic knowledge. Trends Cogn. Sci. 4,
66–72
52 Gelman, S.A. et al. (2002) Children’s use of generics in inductive
inferences. J. Cogn. Devel. 3, 179–199
53 Gelman, S.A. et al. Mother-child conversations about gender: understanding
the acquisition of essentialist beliefs. Monogr. Soc. Res. Child
Dev. (in press)
54 Gelman, S.A. and Raman, L. (2003) Preschool children use linguistic
form class and pragmatic cues to interpret generics. Child Dev. 74,
308–325
55 Pe´rez-Leroux, A.T. et al. (2004) Learning definite determiners:
genericity and definiteness in English and Spanish. Boston University
Conference on Language Development, (Suppl. online: http://www.bu.
edu/linguistics/APPLIED/BUCLD/supp.html)
56 Hollander, M.A. et al. (2002) Children’s interpretation of generic noun
phrases. Dev. Psychol. 38, 883–894
57 Heyman, G.D. and Diesendruck, G. (2002) The Spanish ser/estar
distinction in bilingual children’s reasoning about human psychological
characteristics. Dev. Psychol. 38, 407–417
58 Haslam, N. and Ernst, D. (2002) Essentialist beliefs about mental
disorders. J. Soc. Clin. Psychol. 21, 628–644
59 Strevens, M. (2000) The essentialist aspect of naive theories.
Cognition 74, 149–175
60 Ahn, W. et al. (2001) Why essences are essential in the psychology of
concepts. Cognition 82, 59–69
61 Malt, B.C. (1994) Water is not H2O. Cogn. Psychol. 27, 41–70
62 Sloman,S.A.andMalt,B.C.(2003)Artifactsarenotascribedessences,nor
are they treated as belonging to kinds. Lang. Cogn. Process. 18, 563–582
63 Braisby, N. et al. (1996) Essentialism, word use, and concepts.
Cognition 59, 247–274
64 Braisby, N. (2001) Deference in categorization: evidence for essentialism?
In Proceedings of the 23rd Annual Conference of the Cognitive
Science Society (Moore, J.D. and Stenning, K. eds), Erlbaum
65 Keil, F. (1994) The birth and nurturance of concepts by domains:
the origins of concepts of living things. In Mapping the Mind:
Domain Specificity in Cognition and Culture (Hirschfeld, L.A. and
Gelman, S.A. eds), pp. 234–254, Cambridge University Press
66 Evans, E.M. (2001) Cognitive and contextual factors in the emergence
of diverse belief systems: creation versus evolution. Cogn. Psychol. 42,
217–266
67 Haslam, N. et al. (2002) Are essentialist beliefs associated with
prejudice? Br. J. Soc. Psychol. 41, 87–100
68 Giles, J.W. and Heyman, G.D. (2003) Preschoolers’ beliefs about the
stability of antisocial behavior: Implications for navigating social
challenges. Social Devel. 12, 182–197
69 Yzerbyt, V.Y. et al. (2004) The Psychology of Group Perception: Perceived
Variability, Entitativity, and Essentialism, Psychology Press
70 Nisbett, R.E. et al. (2001) Culture and systems of thought: holistic
versus analytic cognition. Psychol. Rev. 108, 291–310
Review TRENDS in Cognitive Sciences Vol.8 No.9 September 2004 409
www.sciencedirect.com