Generic language and essentialism
Numerous studies with English-speaking children have documented the role
of generic statements, as an important cultural factor, in the
development of essentialism – the belief that categories have internal,
natural causes for, and markers of, category membership (Gelman et al.,
2010; Leshin et al., 2021; Rhodes et al., 2012; 2018). Children learn
about categories, reason about the world, and explain patterns through a
cognitive heuristic to represent categories as “kinds” (Keil, 1989;
Markman; 1989). In fact, children develop this ability early on and are
able to reason about and refer to more abstract groupings of objects,
animals, or people(Sugarman, 1981; Mandler, 2004; Rakison & Oakes,
2003). It is argued that generic language is one of the main mechanisms
through which this learning can occur (Gelman et al., 2010; Rhodes et
al., 2018), although the precise mechanism by which generics early
categorization is still up for debate, an idea that we will return to
later.
In contrast to non-generic expressions that convey information about
specific objects or individuals (“this squirrel sleeps a lot”),
generics transmit information about the kind of objects or individuals
as a group (“squirrels sleep a lot”). Furthermore, while a generic
sentence attributes a common property to a category, in comparison with
quantified expressions (“most squirrels sleep a lot”), generic
statements are silent on how prevalent the property is among members of
the category. Generic language is prevalent in communications with
children, and children understand generic statements by the preschool
age (Gelman, Waxman, & Klienberg, 2008; Gelman, Taylor, & Nguyen,
2004; Cimpian & Markman, 2008).
Perhaps a side-effect of facilitating category representation, generic
statements lead to essentialist beliefs about novel animal species
(Cimpian et al., 2010; Gelman et al., 2010), as well as about familiar
(Cimpian & Markman, 2011; Cimpian & Erickson, 2012) and novel social
categories (Leshin et al., 2021; Rhodes et al., 2012). Psychological
essentialism refers to a set of intuitive, interrelated beliefs about a
category to represent the category as a natural kind and implies that
its members share an underlying essence that causes common observable
features (Gelman, 2003; Rhodes & Mandalaywala, 2017). Essentialist
beliefs about social categories (i.e., social essentialism) have been
observed among a wide age range including children as young as 4 years
and adults across different cultures (Davoodi et al., 2019; del Rio &
Strasser, 2011; Zhu, 2022). Across social category types, social
essentialism is often associated with various kinds of stereotyping and
prejudice among both children (race: Mandalaywala et al., 2019;
ethnicity: Diesendruck & Menahem, 2015) and adults (Mandalaywala et
al., 2017; Haslam & Whelan, 2008). While children readily develop
essentialist beliefs about a number of social categories by early to
middle childhood, only a subset of all possible social categories
becomes the target of essentialist beliefs. That is, among infinite ways
of categorizing people, both children and adults consider only some
categories as informative, immutable, and/or inborn. This raises the
question of what underlying mechanisms lead people to essentialize
particular social categories.
Previous studies suggest that essentialist beliefs about a social
category vary as a function of the social context, including the
structural, cultural, and–of particular interest here– thelinguistic features of the community a child grows up in (Segall
et al., 2015; for a review see Gelman & Fine, 2020). Much of the
research exploring the links between generic language and essentialism
has been done in an experimental context, and with novel social groups,
to ensure that pre-existing beliefs about the group do not interfere
with essentialist beliefs. In seminal work, Rhodes et al. (2012) exposed
children and adults to a novel social category called “Zarpies”.
Sixteen descriptions about Zarpies were presented in a picture book and
in different forms (generic vs. specific) in two within-subjects
experimental conditions. In one condition, descriptions were generic
(“Zarpies love to eat flowers”) and in the other, specific (“this
Zarpie loves to eat flowers”). In the next step, participants answered
questions commonly used to measure essentialist beliefs. Results showed
that when children and adults hear generic, compared to specific
descriptions, they tend to essentialize the category of “Zarpies” more
often. In more recent work, Leshin et al. (2021) found that generic
language (e.g., “Zarpies hop over the puddle”), in comparison to
specific language, increased two critical aspects of essentialist
reasoning. Generic statements lead children to infer that (a)
category-related properties arise from intrinsic causal mechanisms
(e.g., “Zarpies hate to get wet”, rather than “he needed to get to
the other side”) and (b) These features are inflexible (e.g., “he
wouldn’t sometimes choose to splash in it instead”). However, when it
comes to heritability of these traits, they did not find any
difference between generics and specific conditions.
Why (and when) do generics elicit essentialism? Why do generic
sentences lead to social essentialist beliefs? To address this question,
we need to have in mind that essentialism is a multi-faceted construct
(Haslam, Rothschild, and Ernst, 2000). An essentialist view of
ethnicity, such as Turks, for example, implies that (1) Turks are alike
in fundamental and sometimes non-obvious ways (homogeneity), (2)
similarities amongst Turks, and differences between Turks and non-Turks,
are due to some internal reasons as opposed to external factors
(inherence), (3) commonalities between Turks are inborn and often
heritable (naturalness), and (4) being a Turk is immutable, objective,
and discrete and there are no in-between cases – a person is either a
Turk or not a Turk. Whereas these intuitive beliefs are usually
interrelated, it is conceivable that an individual believes only some of
them and not others; this is especially true for children (e.g., Gelman,
Heyman, & Legare, 2007). For example, one can believe that Turks share
many similarities (homogenous ) but also believe that these
similarities are due to some historical, cultural factors, and
non-biological dispositions (and therefore, not innate ).
Additionally, generic statements can be about various types of
properties. They might be about a (pseudo-)biological feature of a
social category (e.g., Turks have strong hearts), or a cultural feature
(e.g., Turks pray every day in the morning). Moreover, generic
statements may or may not be accompanied by explanations describing the
origins of features (e.g., Turks have strong hearts because they live in
the highlands).
Thus, understanding the role of generics in essentialism starts with a
systematic decomposition of the kind of generic statements as they
relate to the different components of essentialism. In the next
sections, we briefly describe two proposed pathways from generics to
different modes of essentialist thinking and summarize related studies
aligned with each account. Note that these accounts are not mutually
exclusive and may work simultaneously in many cases. The main difference
between these two accounts lies in the following question: Does the
linguistic form of generics (i.e., generalizations) in and of
itself and regardless of content elicit essentialist beliefs
about social categories?
Account 1: Explaining generic information through
essentialism – When a property is ascribed to a category through a
generic statement (e.g., Turks eat pastries), the property is viewed as
a non-accidental feature related to this group, thus demanding a
category-based explanation. Based on this account, the psychological
processes that generate the category-based explanations function in a
way that often result in essentialist thinking. For instance, Cimpian
and Salomon (2014) suggest that the mental systems that generate
explanations for observations tend to rely on inherent properties of
objects, instead of structural and relational features (called inherence
heuristics). As a result, to explain the perceived regularities in a
category conveyed through generics, children are likely to assume that
there is an inherent “essence” shared by members of the category which
has caused the mentioned features. For example, Turks eat pastries
because their stomachs were designed to like sweets. This account
suggests that the connection between generics and social essentialism
depends on the function and constraints of our mental
explanation-generating system.
Of note, this account implies that there is no necessary connection
between the linguistic form of generics and essentialization. It is
rather, the content of generic statements that leads to essentialism, in
instances where content conveys biological information. In fact,
children are able to conceive non-essentialist, structural explanations
for regularities in a category as a result of exposure to generic
statements that emphasize the environment (Vasilyeva & Lombrozo, 2020).
Based on this account, if information contained in or accompanied by
generics signals that referenced regularities are due to factors
external to the category, generics do not necessarily elicit
essentialist beliefs (Noyes and Keil, 2020; Hoicka et al., 2021). For
example, Noyes and Keil (2020) found that by hearing generic sentences
about a novel social category (“Vawnsies”), children and adults
expressed essentialist beliefs (specifically inheritability) onlywhen generic sentences had biological content (“Vawnsies feel sick when
they drink milk”) and not when they were cultural (“Vawnsies believe
that fish talk to God”). Thus, the contextual information attached to
the generics matter in whether or not generic language leads to
essentialism (see also Hoicka et al., 2021).
Account 2: Confirming essences through generic
information – On the second account, generics do not lead to
essentialism, but only confirm essentialist reasoning. On this
account, the pathway from generics to social essentialism is independent
of the content of the generic statements and the consequent inherent
explanations. According to this account, children learn to essentialize
categories very early on and generic language signals to them which
social categories are salient in their communities and prime
“candidates” for essential reasoning (Benitez et al., 2022,
Foster-Hanson et al., 2022). Consequently, regardless of the explicit
information communicated in a generic sentence, using generic terms
(e.g., “Turks”) signals to children that these categories are
culturally relevant and appropriate targets for essentialization. In
support of this hypothesis, researchers have found that generic
sentences can elicit more essentialist beliefs than specifics, even when
they get negated or get corrected by a knowledgeable character
(Foster-Hanson et al., 2022; Foster-Hanson et al., 2016). Thus, on this
account, one mechanism through which generics contribute to the
development of essentialism is through the linguistic form of generic
statements, which signals to children the relevant ways to categorize
the social world.
To sum up, the second account suggests that generic sentences,
regardless of their content, elicit generalization and internal
explanations for the behaviors and preferences of category members. The
first account, however, posits that only through additional biological
information in their content, generic statements lead to essentialist
reasoning, especially beliefs about heritability.