bs_bs_banner

DOI: 10.1111/ejop.12033

What Is Social Construction?

E. Diaz-Leon

Abstract: In this paper I discuss the question of what it means to say that a
property is socially constructed. I focus on an influential project that many social
constructivists are engaged in, namely, arguing against the inevitability of a trait,
and I examine several recent characterizations of social construction, with the aim
of assessing which one is more suited to the task.

1. Introduction

It is common place in the humanities and the social sciences to claim that certain
human features, such as someone’s gender, race, or sexual orientation, are
socially constructed. This view about the nature of these human categories
(known as Social Constructionism) is supposed to be in contrast with two rival
views, namely, Biological Realism (that is, the view that a certain category is a
biologically real kind), and Anti-Realism (that is, the view that a certain category
is empty: nothing belongs to it, the corresponding expression does not refer to
anything). Social constructionism is taken to be a realist account of the nature of
a certain category: it is claimed that the category is a real feature of human
beings, but it is determined by social, rather than natural or biological properties.
In this paper, I will focus on the question of what it means to say that a
category is socially constructed. As we will see, it does not make much sense to
look for the notion of social construction, because the label can be, and has been,
used in different ways. Rather, a better strategy seems to be to focus on some
specific projects and aims that social constructionists may have in mind, and ask,
with respect to each project, which notion of social construction is most useful.
In particular, in this paper I will focus on a very influential project that social
constructionists are typically engaged in, namely, arguing against the inevitability
of a trait, and we will discuss which notion of social construction is most useful
in this case. More in particular, we will examine some recent characterizations of
social construction, and we will ask, for each of them, whether it would entail
that a certain human feature is not inevitable in the required sense. In addition,
we will explore the related questions of whether the notions of social construc-
tion discussed here entail that a property is not intrinsic, or not biological. In
short, we are interested in clarifying the notion of social construction because we
are interested, first, in clarifying what follows from the claim that a certain
category is socially constructed, and second, in clarifying what kind of evidence
could be used in order to establish that a certain human feature is socially
constructed.

European Journal of Philosophy 23:4 ISSN 0966-8373 pp. 1137–1152 © 2013 John Wiley & Sons Ltd
1138 E. Diaz-Leon

2. Social Construction, Contingency, and Inevitability

Ian Hacking (1999) has done more than anyone else in trying to clarify what is
going on in different social constructionist projects and what notion of social
construction is at issue. He claims that many social constructionists about
different categories are interested in the following project:
Social constructionists about X tend to hold that:
(1) X need not have existed, or need not be at all as it is. X, or X as it
is as present, is not determined by the nature of things; it is not
inevitable.
Very often they go further, and urge that:
(2) X is quite bad as it is.
(3) We would be much better off if X were done away with, or at least
radically transformed. (Hacking 1999: 6)
This is what I call the project of arguing against the inevitability of a trait. The
project is especially appealing when (2) and (3) are held about a certain category:
in those cases, claim (1) opens a path for change and transformation regarding
what is seen as a harmful state of affairs. But what does (1) mean exactly? Or
more precisely: how should (1) be understood, so that claims to that effect really
open a path for change and transformation?
A preliminary clarification is in order: (1) cannot just be saying that X is
contingent. If the point of social constructionism were just to say that a certain
feature X is contingent (that is, that it might not have obtained) and is in some
sense ‘not inevitable’, then many properties that are presently instantiated in the
world, and indeed, most biological properties, would be contingent in that
sense.1 If social constructionism were to be understood along these lines, that is,
in terms of the contingency of the instantiation or the current distribution of a
certain property, then social constructionism about X would not be incompatible
with biological realism about X. But as we said at the beginning, this is one of
the desiderata for a notion of social construction: social constructionism and
biological realism about X seem to be competing views, at least in principle. In
addition, and more relevantly for our purposes: if we understand (1) merely in
terms of what is metaphysically possible (that is, in terms of what is the case
in some metaphysically possible world), then this claim doesn’t really open the
way for a feasible strategy to achieve social change and social justice. If
the instantiation or distribution of X in the world is not inevitable merely in the
sense that, if the laws of nature were different, X would not have existed, or
would have been different from the way it actually is, then there isn’t much we
can do about the current instantiation or distribution of feature X, assuming that
changing the laws of nature is not really in our hands anyway.
Therefore, the point of (1) is rather to claim that the instantiation or distri-
bution of X is contingent upon certain social events and social arrangements: if
those social events and arrangements were different, then facts about X could be

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1139

different. As Hacking puts it, the point of (1) is to claim that ‘X was brought into
existence or shaped by social events, forces, history, all of which could well have
been different’ (Hacking 1999: 7). In this way, when (1), (2) and (3) are all
endorsed, with respect to a certain feature X, (1) provides a method for getting
rid of X, to wit: by changing the social arrangements that give rise to X.2
The question I want to examine in this paper is this: which notion of social
construction is more adequate for the project of arguing against the inevitability
of a trait? In order to explore this question, I will focus on several characteri-
zations of social construction in the recent literature, and I will ask whether they
entail (1) or not. The aim of this inquiry is not just to come up with an artificial
notion of social construction that will give us the results that we want; rather,
what we are looking for is a method for assessing claims and arguments of the
following form: a certain human category, say, race, or gender, or sexual
orientation, has such and such features, and therefore it is socially constructed,
and therefore it is not inevitable, and therefore facts about the instantiation and
distribution of this category (which are taken to be unjust) can be transformed
by means of social action. My main aim is to examine arguments of this form,
which are obviously important for social and political reasons, so as to tell apart
those versions of the argument that work from those that don’t work.

3. The Construction of Ideas

In the recent literature, several notions of social construction have been intro-
duced, and some important distinctions have been made. The first crucial
distinction we have to address concerns the objects of the construction: what is
constructed? Hacking (1999) distinguishes two notions of social construction
along these lines: the social construction of ideas (or concepts, or theories, that is,
mental representations), and the social construction of objects (or individuals,
properties, kinds, facts, that is, entities in the world, as opposed to our mental
representations of them).3
We will focus on the construction of ideas first. Idea-constructionism about a
certain category X, then, amounts to the claim that our ideas, concepts and
theories about X are the result of contingent social factors. As Sally Haslanger
(2003) argues, this is not a very controversial claim, and on some formulations,
it might be trivially true: it is obvious that the theories and concepts we actually
have about any domain are to a certain extent the result of contingent historical
and social events. For instance, we need individuals with certain psychological
traits in order to produce the concepts and theories that we have; furthermore,
technological developments, historical events, and even social and political
factors (such as what kind of research gets funded) all have an input into what
kind of theories and concepts are used today and the level of understanding we
have in the different areas of human knowledge, but this does not entail that our
theories and concepts, nor the entities they are about, are socially constructed, in
any interesting or controversial sense. Of course, this (partial) dependence upon

© 2013 John Wiley & Sons Ltd

1140 E. Diaz-Leon

social factors will also hold with respect to theories in the natural sciences, and
in particular in the biological sciences, so then the claim of idea-constructionism
about X won’t really be incompatible with biological realism about X either.
In order to make the claim of idea-constructionism more substantial,
Haslanger, following Hacking, develops the claim into three further sub-claims,
as follows: ‘[idea-]constructionists with respect to a domain D . . . are sympa-
thetic to (a) the contingency of our understanding of D; (b) nominalism about
kinds in D, or more precisely, a denial that the domain D has an inherent
structure; and (c) an explanation of the stability of our understanding of D in
external rather that internal terms’ (Haslanger 2003: 304). As I understand it, an
explanation of the stability of a theory in internal terms consists in explaining
why we currently hold a certain theory in terms of factors that are relevant to
the justification and rationality of the theory, such as the degree of evidence that
we have, the internal coherence of the theory and other theoretical virtues such
as simplicity, elegance, and so on; whereas we should understand explanations
in external terms as those that appeal to factors that do not necessarily make the
theory more rational or justified, such as social and political factors that could
affect our disposition to hold a certain theory rather than another. Once we
understand idea-constructionism in terms of those three claims, we could then
formulate a possible debate between idea-constructionism about a certain category
X, on the one hand, which would claim that our understanding of X depends
entirely on social factors, not on facts about X, and idea-determinism about X, on
the other hand, which would claim that our understanding of X depends entirely
on the way things are with respect to X, but does not depend on social factors
at all. Haslanger argues that both views, so characterized, seem trivially false: for
any domain, whether social or natural, our theories and concepts will depend in
part on the way things really are, and in part on social and historical factors
underlying our understanding of that domain. According to Haslanger, this
debate doesn’t really capture what is at issue when scholars about gender, race
or sexual orientation claim that these human features are socially constructed.
I pretty much agree with Haslanger’s analysis of Hacking’s discussion about
the social construction of ideas: as she puts it, ‘plausibly our ideas and classi-
fications are the product of some combination of worldly input from perception
and experience and social input from languages, practices, and the like. The
debate as presented by Hacking is not very interesting because neither extreme
view is plausible and very little is offered to cover the more interesting
middle-ground’ (Haslanger 2003: 307). This remark, though, gives rise to an
interesting question: is there any way of formulating the debate between
idea-constructionism and idea-determinism which covers that middle-ground?
As Haslanger herself points out, it seems clear that discussions about whether (a)
and (c) are true, for some concept or theory, are not very interesting, as we have
seen. However, claim (b) above seems to be a more substantial claim, so perhaps
this could give rise to a more substantial debate about whether nominalism
about a certain concept X is true or false. This debate could be formulated
(roughly) as follows: are facts about the instantiation of X mind-independent or

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1141

not? That is to say, is the property that all things that fall under X have in
common just a matter of being classified in that way by individuals like us, or
do they share an underlying nature or ‘essence’, independently of how we
classify things? This debate seems more interesting than the question of whether
our understanding of X depends entirely on social factors, or entirely on the way
things are.4 However, the debate about nominalism is a metaphysical debate about
the nature of the kinds and properties that we are trying to understand, not a
debate about the nature of our concepts and theories about those kinds and
properties, as the label ‘idea-constructionism’ and ‘idea-determinism’ might
suggest. In fact, as Haslanger also emphasizes, this metaphysical debate is
strongly related to the discussion concerning the construction of objects and
kinds, rather than idea-construction, which we will examine next.

4. The Construction of Objects and Kinds

We will now turn to the social construction of objects. The idea here is that some
social agents or social factors produce and control some individuals and/or the
properties that they have. How should we understand this intriguing claim?
Haslanger (2003) and Mallon (2008) have both emphasized that we should make
the following important distinction, with respect to which kind of construction
process is at issue: we should distinguish between causal social construction and
constitutive social construction, as follows:
Causal Social Construction:
X is socially constructed causally as an F iff social factors (i.e. X’s
participation in a social matrix) play a significant role in causing X
to have those features by virtue of which it counts as an F.
(Haslanger 2003: 317)
X causally constructs Y if and only if X causes Y to exist or to persist
or X controls the kind-typical properties of Y. (Mallon 2008: 5)
Constitutive Social Construction:
X is socially constructed constitutively as an F iff X is of a kind or sort
F such that in defining what it is to be F we must make reference to
social factors (or: such that in order for X to be F, X must exist within
a social matrix that constitutes Fs). (Haslanger 2003: 318)
X constitutively constructs Y if and only if X’s conceptual or social
activity regarding an individual y is metaphysically necessary for y
to be a Y. (Mallon 2008: 6)
The distinction here should be clear: An object or kind is causally socially
constructed when social factors or social agents are causally responsible for the
existence of the object or the instantiation of the corresponding properties. On
the other hand, an individual or property F is constitutively socially constructed

© 2013 John Wiley & Sons Ltd

1142 E. Diaz-Leon

when it is part of the definition of what it is for someone to be an F, or part of

the nature of being an F (i.e. what makes someone an F), that Fs stand in some
relation to social agents or social factors.5
It will be useful to consider some examples of things that fall under one kind
of social construction but not the other. For instance, Mallon claims that
artifactual objects are arguably causally but not constitutively socially con-
structed: ‘Representations expressing the concept watch are causally necessary for
some materials to become a watch, but they are not metaphysically necessary. It
is metaphysically possible, however unlikely, that we could walk across a heath
and find a watch that had always been there’ (Mallon 2008: 6). That is to say, in
order for there to be any watches, we need individuals with certain intentions
and certain abilities to engage in certain practices, in order to produce objects that
can satisfy the function of a watch. Many centuries ago, before there were any
human beings on earth, there were no watches, so watches are causally socially
constructed by human beings. However, according to Mallon, it is not part of the
definition (nor is it part of the nature) of what it takes for something to be a
watch, that something is a watch only if it bears a certain relation to human
beings. In this sense, watches are mind-independent objects: it is metaphysically
possible that we could find an object in nature that is identical to a watch, which
has not been created by humans (even if this is extremely unlikely, and perhaps
even nomically impossible, that is, impossible given the laws of nature in the actual
world). Therefore, watches are not constitutively socially constructed.6
What kind of things will be constitutively socially constructed, then, if artifacts
and other human creations are not? A good example is that of institutional or social
roles, such as being a University Professor, or a landlord, or a tenant. Haslanger
presents the following example: ‘Consider, for example, what must be the case for
someone to be a husband in the contemporary United States: A husband is a man
legally married to a woman. Being a man legally married to a woman does not
cause one to be a husband; it is just what being a husband consists in’ (Haslanger
2003: 317). The property of being a husband is clearly socially constructed in the
constitutive sense: it is part of what it means (and what it is) to be a husband, that
there must exist certain legal arrangements and certain social practices, and one
must stand in certain relations with respect to these social practices in order to be
a husband. Without those legal arrangements and social practices (or some
equivalent ones), no one would be anyone’s husband, so this human kind is
constitutively connected to those social and legal practices.
Paul Boghossian (2006) has also provided a similar distinction between
causal and constitutive notions of social construction, although the characteri-
zation that he offers seems to be a combination of both, which in my view is
misleading. He says: ‘to say that [something] was socially constructed is to [say]
that it was built by a society, by a group of people organized in a particular
way, with particular values, interests and needs’ (Boghossian 2006: 16). This
characterization is very similar to the notion of causal social construction above,
but Boghossian makes clear that it is not the relevant notion in order to
characterize social constructionism: ‘our social construction theorist is not inter-

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1143

ested in cases where, as a matter of contingent fact, some fact is brought into
being by the intentional activities of persons, but only in cases where such facts
could only have been brought into being in that way. In the intended technical
sense . . . it has to be constitutive of a given fact that it was created by a society
if it is to be called “socially constructed”. For example . . . a piece of paper’s
being money is a socially constructed fact in the technical sense, for it is
necessarily true that it could only have come to be money by being used in
certain ways by human beings organized as a social group’ (2006: 17). That is,
Boghossian is defining constitutive social construction in terms of properties
that are necessarily caused by the intentional activities of human beings. The
difference between this characterization and our working characterization
above is that according to Haslanger’s and Mallon’s definitions, constitutive
social construction of P does not require that the relation between social agents
and property P be a causal relation (e.g. as Haslanger explains, what makes
someone a husband is that he is married to someone, but this does not cause
someone to be a husband). Boghossian on the other hand understands socially
constructed properties as those that could only be brought into being as the
causal result of human agents. Is this a useful notion? I think that given certain
assumptions, Boghossian’s notion and our notion above are not co-extensive.
For instance, if we assumed that nomically impossible worlds are not metaphysi-
cally possible (that is, that there are no metaphysically possible worlds with
different laws of nature), then many causally socially constructed properties
would also count as constitutively socially constructed according to
Boghossian’s notion, but not according to our notion above. For instance, let’s
consider the following property of a 1-dollar coin: being a piece of metal with
such and such physical properties (shape, size, weight, and so on). Typically,
something comes to have this property only as a result of intentional activities.
We could also assume that, given our laws of nature, something could come
to have those properties only as a result of intentional activities. If these laws
of nature turn out to be metaphysically necessary, then it would be metaphysi-
cally necessary in order for something to have those physical properties, that
it be created by individuals organized in certain ways. But still, there seems to
be an intuitive difference between those physical properties of a 1-dollar coin,
which are the causal result of human activities, and the clearly constitutively
socially constructed property of being a 1-dollar coin, which is also instantiated
by that piece of metal. The property of being a 1-dollar coin is socially
constructed in a stronger sense than the property of being a piece of metal with
such and such physical features: whereas we can grant for the sake of the
argument that the latter can be brought into being (necessarily) only by means
of human interaction, the former depends on human interaction in a stronger
sense: it is part of the nature of money that something counts as money only
when it is used in certain ways by intentional beings. Boghossian’s characteri-
zation of constitutive social construction in terms of necessary causal connec-
tions does not capture this distinction, and is therefore inadequate for our
purposes here.

© 2013 John Wiley & Sons Ltd

1144 E. Diaz-Leon

The distinction between causal and constitutive social construction is extremely

important in order to make sense of the different social constructionist projects
that have been pursued and their different aims. In the remaining of this paper,
I want to point out three important consequences of this distinction. First, I will
argue that it makes a significant difference with respect to the project of arguing
against the inevitability of a trait; second, I will argue that it also makes a crucial
difference with respect to the project of arguing that a certain trait is relational
rather than intrinsic; and finally, I will argue it makes a difference regarding the
project of arguing that a certain trait is not biologically real.

5. Social Construction and the Possibility of Social Change

Our first question, then, will be to consider whether social construction in either
of these two senses entails that a certain trait is not inevitable, and can be
transformed by means of social action. Let’s examine the notion of causal social
construction first. As we have seen, a property is causally socially constructed
when social agents and social factors play a causal role in bringing about, or
controlling, the instantiation of such a property. Does this mean that if a property
is socially constructed in this sense, then we can change facts about the
instantiation and distribution of the property just by changing our social
arrangements and practices? Not necessarily: if a property P is causally but not
constitutively constructed, it might be very hard to change facts with respect to
P just by changing our present social practices. For instance, Haslanger discusses
the following example of a property that is causally but not constitutively
constructed: ‘some feminists claim that some anatomical phenomena have social
causes, for example, that height and strength differences between the sexes are
caused by a long history of gender norms concerning food and exercise’
(Haslanger 2003: 317). According to this claim, traits such as having a certain
height or a certain strength have been caused in part by social factors, such as
cultural norms regarding which food and activities are appropriate for women,
and therefore they would be causally socially constructed, but they are not
constitutively socially constructed because it is not part of what having a certain
height or a certain strength consists in that it must be the result of certain
cultural and social practices. (That is, we could find individuals with exactly the
same strength and the same height, as a result of purely natural or biological
factors.) If this claim about the causal origin of these traits in women is correct,
then we would have a clear case in which some gendered traits are causally (but
not constitutively) socially constructed, but this does not necessarily mean that
they can be changed or transformed just by means of social action. In particular,
the fact that a certain individual has a trait that has been causally socially
constructed in this way, does not automatically pave the way for creating feasible
social strategies for changing those traits in those individuals. Perhaps uncover-
ing the social origins of these traits will give rise to strategies for preventing the
instantiation of those traits in future generations, but this will not necessarily
help to change the facts that we take to be unjust today.

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1145

However, the situation is very different with respect to constitutive social

constructions. If we claim that a certain gendered trait is constitutively socially
constructed (for instance, being a wife, or being a housewife, or even being a
woman, according to Haslanger’s own definition, according to which being a
woman is a matter of occupying a certain position of subordination7), then this
claim will immediately open a way for creating feasible strategies for social
change. For example, if it is part of the definition of being a woman that one
occupies a certain position of subordination with respect to a hierarchical system
in which some individuals are privileged and some individuals are subordi-
nated, then if we change those social practices of privilege and subordination,
there will no longer be any women (and likewise for wives, or husbands, or
housewives, or university professors).
This point is especially important with respect to Hacking’s discussion of
social constructionist projects, which he reinterprets as searching for a way to
get rid of certain facts and traits that are considered unjust. But, paradoxically,
the kind of social construction that Hacking focuses on is causal social con-
struction. This, I believe, is a mistake. Haslanger makes a similar point:
‘Hacking is right that the goal is often to challenge the appearance of inevi-
tability of the category, to suggest that if social conditions are changed, it
would be possible to do away with the category. But an important first step is
to make the category visible as a social category’ (Haslanger 2003: 318–9). More
precisely, we have to make clear that the category is constitutively socially
constructed. The main motivation for this claim, in my view, is the following:
if we want to suggest a method for doing away with the category by means
of changing social conditions, merely claiming that a certain trait is causally
socially constructed does not have the same implications as the claim that the
trait is also constitutively constructed, as we have seen in the case of women’s
height and strength above. Talking about the causal origins of a certain trait
does not necessarily open a way for changing current conditions. But as
Haslanger points out, showing that a category is constitutively socially con-
structed, that is, revealing the constitutive connections between instantiating a
certain category and standing in a certain relation to certain social practices,
opens a clear path for social change: just change those social practices, and
social change will automatically follow.
Of course, how easy or how difficult it is to actually change those social
practices and achieve social change will depend on the kind of social practices
that determine the socially constructed property in question. In some cases,
changing the distribution of some causally socially constructed properties might
be easier, from a practical point of view, than changing the distribution of some
constitutively socially constructed properties. For instance, it might be relatively
easy to prevent the instantiation of the causally socially constructed property of
wearing make-up (e.g. by changing its most common social causes, namely,
norms of feminine appearance), but it might be very hard to change the social
practices that determine the economic differences between men and women
(which are constitutively socially constructed). However, I think the distinction

© 2013 John Wiley & Sons Ltd

1146 E. Diaz-Leon

between causal and constitutive social construction is still significant from a

conceptual point of view, because it is still the case that changing the social
factors that typically cause the instantiation of some property will not automati-
cally change the distribution of such property. For example, changing our
cultural gendered norms regarding feminine appearance might not be enough to
change the fact that many women wear make-up, because other cultural norms
might also operate; and moreover in the case of some other causally socially
constructed properties, such as having plastic surgery, the causal effects are more
permanent. On the other hand, it seems clear that in the case of constitutively
socially constructed properties, changing the social factors that are metaphysi-
cally (or conceptually) necessary for the instantiation of a property will auto-
matically entail (with conceptual and/or metaphysical necessity) that the
corresponding property is no longer instantiated, even if it is very difficult to do
so, from a practical point of view.
Hacking has offered a very helpful survey of the different projects that a social
constructionist might have in mind (Hacking 1999: 19–20). All these projects start
by assuming what he calls historical constructionism: ‘X is the contingent result of
historical events and forces, therefore (1): X need not have existed, is not
determined by the nature of things, etc.’; but they differ with regards to how
they react to that claim, to wit: ‘at this stage we cannot help but treat X as “part
of the universe”, but our way of thinking may evolve so that X is no longer
viewed in this way’ (ironic constructionist); ‘although we cannot at this stage see
how to avoid X, we should try to improve it’ (reformist constructionist); ‘if we
understand the function of X socially, we will see that it should have no appeal
for or authority over us’ (unmasking constructionist); ‘we would be much better
off if X were done away with or radically transformed’ (rebellious constructionist);
and finally ‘we would be much better off if X were done away with or radically
transformed. In addition, the revolutionary constructionist acts to do away with
X’ (revolutionary constructionist). However, as I have argued, these goals of social
change are at odds with Hacking’s own understanding of social construction as
a causal process. As Haslanger puts it: ‘Because on Hacking’s view social
constructionisms are concerned with causal claims, it doesn’t capture what’s
interesting in claiming that something is, perhaps surprisingly, a social kind’
(Haslanger 2003: 318). Haslanger’s main reason for saying this seems to be that
causal claims of that form do not seem to capture what social constructionists
actually have in mind when they claim that a certain trait is socially constructed:
‘social constructionists are often not interested in the causes of our ideas and the
social forces at work on objects, but are interested in how best to understand a
given kind, and in particular whether it is a natural or social kind’ (Haslanger
2003: 318). Given the considerations so far, now we can see why this is what
social constructionists should be doing anyway, if they want to pursue the
different constructionist projects that Hacking described. For the sort of claim
that will really make a difference, with respect to all those projects above, is the
claim that a certain trait is socially constructed, not merely causally but also
constitutively so.8

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1147

6. Social Construction and Intrinsic vs. Relational Properties

The second issue I want to explore, which is closely connected to the previous
one, is the question of whether the social construction of a human kind entails
that it is not an intrinsic feature of human beings. For example, in a very
influential essay about the social construction of sexual orientations, Edward
Stein characterizes the debate between social constructionism and what he calls
‘essentialism’ regarding sexual orientations (which we can understand in terms
of biological realism, for our purposes here) as follows: ‘Essentialists hold that
a person’s sexual orientation is a culture-independent, objective and intrinsic
property while social constructionists think it is culturally-dependent, relational
and perhaps, not objective. (Roughly . . . an intrinsic property is one that a
person has non-relationally, i.e. “inside” him or her; in other words, an intrinsic
property is one that a person could have even if she were the only person or
thing in the world . . . like having a certain blood type or being a person taller
than six feet)’ (Stein 1992: 325–6). For our purposes here, I will understand
‘intrinsic property’ along these lines: a property P of an individual x is an
intrinsic property if and only if any duplicate of x in any possible world would
also have property P, including a duplicate of x that is the only existing entity
in that possible world. Given this definition, it is clear that social construction in
the causal sense does not entail that a property is not intrinsic, for there are
properties that are causally socially constructed but are intrinsic properties of an
individual. For example, the property of having a tattoo is causally socially
constructed because in our world, social factors play a crucial causal role in
bringing about this property (e.g. we need certain technologies and certain social
practices in order to get individuals with tattoos), but having a tattoo is
nonetheless an intrinsic property of an individual, not a relational one: If I have
a tattoo, any physical duplicate of me will also have an exactly identical tattoo.9
Another example is the property of being circumcised: we can assume that for
many circumcised individuals in the actual word, social factors and social agents
play a crucial causal role, but nonetheless, any physical duplicate of a circum-
cised individual would also be circumcised, no matter what the different causal
origins are in each particular possible world. Hence, this property is also intrinsic
but causally socially constructed. Finally, and combining Stein’s own example
and Haslanger’s example above, the property of having a certain height, which
according to some feminists many women have as a result of certain social
practices, is causally socially constructed according to this view, but is clearly
intrinsic.
On the other hand, any property that is constitutively socially constructed will
be relational, not intrinsic, given that by definition, what makes a property
constitutively socially constructed is the fact that something satisfies the property
only if it bears a relation to certain social practices and communities. Thus, if we
change or remove these social practices and communities, the individual will no
longer have the corresponding trait, and therefore it is not true that any physical
duplicate of the individual will have the property, so the property fails to be

© 2013 John Wiley & Sons Ltd

1148 E. Diaz-Leon

intrinsic. This is strongly connected to the point I made above, namely, that the
kind of social construction that is more closely relevant to the goal of achieving
social justice by means of social action is the constitutive social construction of
a certain human kind, because this claim will have clearer social and political
significance concerning the feasibility of different social strategies for achieving
social change.
In any case, it is clear that Stein’s characterization of social constructionism in
terms of a property being relational is misguided: not all versions of social
constructionism entail that a property is relational rather than intrinsic. Many
social constructionist projects, such as the feminist project of showing that some
gendered traits such as women’s strength and height have a social origin rather
than a biological one, do not necessarily entail that those traits are not intrinsic.

7. Social Construction and Biological Realism

Finally, I want to consider to what extent these notions of social construction

help to make the distinction between social constructionism and biological
realism clearer. It is clear that if a property is constitutively socially constructed
then it is not a (paradigmatic) biological property, because it is part of the nature
of the property that someone must stand in certain relations to certain social
practices and arrangements, in order to instantiate the property. However, it is
not so clear what happens in the case of causally socially constructed properties.
For example, in the case of the anatomical differences between men and women
described above, where it is suggested that this is due in part to social factors,
these properties would be causally socially constructed but they still seem to be
biological properties in some sense. Another interesting example is suggested by
Philip Kitcher’s account of race: according to Kitcher (1999), races are (roughly)
breeding populations that have been reproductively isolated, mostly as a result
of social factors. If this theory about the nature of human races is correct, then
it follows that races are causally socially constructed. But they can still be seen
as biologically real properties, at least insofar as one wants to consider
reproductively isolated breeding populations as biologically real kinds (for
instance, Robin Andreasen (1998), who endorses a similar account of races
according to which they are also reproductively isolated breeding populations,
clearly classifies this account as a version of biological realism about races).
Given these considerations, we can conclude that social constructionism regarding
X does not necessarily entail that biological realism about X is false: only when we
specify that X has been constitutively socially constructed, can we infer that X is
not biologically real.

8. Conclusion

In this paper I have examined several recent characterizations of social con-

struction, and I have argued that the notion of the social construction of kinds

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1149

and properties (as opposed to the social construction of concepts and theories),
in the constitutive sense (as opposed to the merely causal sense), is the most
useful notion in order to capture an important aim of many social
constructionist projects, namely, to show that the current instantiation and/or
distribution of a certain property, which is taken to be unjust, is contingent
upon certain social practices, which should be changed if we want to achieve
social justice. Moreover, I have clarified some further different consequences of
these two versions of social constructionism, namely, causal vs. constitutive
social constructionism about X. As we have seen, only the latter, but not the
former, implies that X is not intrinsic, or not (paradigmatically) biological. This
result can help us to evaluate arguments to the effect that a certain property
or trait is (or is not) socially constructed, because from the claim that a certain
trait is intrinsic or biologically real we can only infer that it is not socially
constructed in the constitutive sense, but we cannot rule out that it is causally
socially constructed.10

E. Diaz-Leon
Department of Philosophy
University of Manitoba
Winnipeg, MB, Canada R3T 2N2
[email protected]

NOTES
1
One way of showing that biological properties, as currently instantiated, are
contingent is to assume that the laws of nature are contingent.
2
Of course, when the social factors that give rise to feature X involve crucial appeal
to historical facts and past events, then it will not be easy to get rid of X, since we cannot
really change the past. However, claims to this effect could still open a path for preventing
the instantiation of X in the future, and in this way provide a strategy for achieving social
change as well.
3
See also Mallon (2007, 2008) for further discussion of this distinction.
4
It should also be noted that, even if we agree that these two extreme views seem
misguided, this still leaves room for an interesting discussion about to what extent social
factors influence our choices of theories.
5
A couple of clarifications are in order. First, according to Haslanger’s characteriza-
tion of a property that is causally socially constructed, X’s instantiating F is socially
constructed when social factors play a significant role in causing X to have those features
by virtue of which it counts as an F. There are at least two ways of understanding this idea:
first, we could understand this as requiring that social factors play a role in determining
what counts as an F (i.e. how we use the concept of F); second, we could understand this
as merely requiring that social factors play a role in causing X to instantiate either F itself,
or some properties that determine F. In my view, it is this second, metaphysical reading
the one that is intended here: the former would just be a way of going back to the
construction of ideas rather than objects or kinds. Second, there is a subtle but significant
difference in the characterizations of constitutive social construction by Haslanger and
Mallon, respectively: whereas Haslanger talks about those factors that we must make

© 2013 John Wiley & Sons Ltd

1150 E. Diaz-Leon

reference to in defining what it is to be F, Mallon talks about those social factors that are
metaphysically necessary for something to instantiate the corresponding property. As I
understand it, Haslanger is focusing on those social factors that are part of the definition
of an F, or in other words, conceptually necessary for something to instantiate F, whereas
Mallon is focusing on those social factors that are part of the nature of an F, or in other
words, metaphysically necessary for something to instantiate F. It is important to notice this
difference because in some cases, the conditions that are conceptually necessary for
something to be F, and the conditions that are metaphysically necessary for something to
be F, might come apart. For instance, it is widely accepted that it is metaphysically
necessary for something to be water that it be composed of H2O, but it is not conceptually
necessary. (And a bit more controversially, some philosophers accept that it is conceptu-
ally necessary for something to be water that it be watery stuff, but it is not metaphysi-
cally necessary.) However, for our purposes we can put this distinction aside and treat
both characterizations as more or less equivalent, since we can assume for the sake of
discussion that social kind terms are not “semantically unstable” in the way natural kind
terms are.
6
It could be argued that it is part of the nature of any artifact object that it must have
been created with certain intentions. If so, it would be metaphysically necessary that if x
is a watch, it must bear a certain relation R to human beings or some other individuals
with certain psychological properties (namely, the relation of being created by certain
individuals with certain intentions). If this intentional account of artifacts is accepted, then
it will follow that artifacts are also constitutively socially constructed. But on an alternative
account of artifacts (which we can call the functional account), according to which what
it takes for something to be (say) a watch is that it satisfies a certain function, that would
not follow, since it will be metaphysically possible for something to satisfy the corre-
sponding function without having been created by human beings (even if this is perhaps
nomically impossible). It is important to notice, though, that even if the functional account
of artifacts is rejected (and the intentional account is endorsed), we would still have an
example of a property that is causally but not constitutively socially constructed: just take
the functional property of a watch that according to the functional account characterizes
watches (even if you reject that this is what makes a watch a watch), and this property
(or at least, their instantiations in the actual world) will be causally but not constitutively
socially constructed.
7
Haslanger’s characterization of woman goes as follows: ‘S is a woman iff (i) S is
regularly and for the most part observed or imagined to have certain bodily features
presumed to be evidence of a female’s biological role in reproduction; (ii) that S has these
features marks S within the dominant ideology of S’s society as someone who ought to
occupy certain kinds of social position that are in fact subordinate (and so motivates and
justifies S’s occupying such a position); and (iii) the fact that S satisfies (i) and (ii) plays
a role in S’s systematic subordination, i.e., along some dimension, S’s social position is
oppressive, and S’s satisfying (i) and (ii) plays a role in that dimension of subordination’
(2003: 319).
8
Interestingly enough, Haslanger’s more recent characterization of social construction
does not focus so much on this constitutive element, but rather on the (non-)naturalness
of the property that unifies instances of the kind: ‘One interpretation of [the claim that
categories once assumed to be “natural” are in fact “social” or, in the familiar lingo,
“socially constructed”] is that although it is typically thought that what unifies the
instances of such categories is some set of natural or physical properties, instead their
unity rests on social features of the items in question. Social constructionists pursuing this

© 2013 John Wiley & Sons Ltd

 What Is Social Construction? 1151

strategy . . . aim to “debunk” the ordinary assumption that the categories are natural, by
revealing the more accurate social basis of the classification’ (2006: 89). However, it is
important to realize that this characterization, in itself, does not distinguish between
causal and constitutive social construction. It is true that all constitutively socially
constructed properties, according to our characterization here, will be social rather than
natural properties. But it is not true that, if a property is not natural, then it is constitutively
socially constructed. There are many properties that we wouldn’t classify as natural (for
instance, the artifactual properties discussed above, such as being a watch), but as we
have seen, these are causally but not constitutively socially constructed.
9
As before, someone could reply that the property of having a tattoo necessarily
involves a certain relation to the intentions of the creators of the tattoo, and because of this,
not every physical duplicate of me would have the same tattoo, even if they have the same
physical properties and the same amounts of ink on their skin with exactly the same shapes
and so on. In response, we could say that in this case the property that is causally socially
constructed but not relational is the property of having such and such physical marks on
my skin, forming such and such figures, which any duplicate of me would also share.
10
I am grateful to the University of Manitoba Internal Research Grant Programme,
and the Social Sciences and Humanities Research Council Standard Research Grant
Programme, for financial support. I have presented ancestors of this material at several
venues, including: a panel of the Eastern Society for Women in Philosophy at the Eastern
APA in New York, a panel of the Society for Analytic Feminism at the Central APA in
Minneapolis, the British Society for the Philosophy of Science conference at the University
of Sussex, the congress of Logic, Methodology and Philosophy of Science at Nancy
(France), the Canadian Society for Women in Philosophy conference at Mount Royal
University in Calgary, and colloquiums at the University of Manitoba Institute for the
Humanities, and the University of Murcia (Spain). In addition, this paper was the target
of in-depth discussions at the Mentoring Workshop for Women in Philosophy at the
University of Massachusetts at Amherst, and the PERSP Metaphysics Seminar at the
University of Barcelona. I am very grateful to the audiences in all those occasions for very
useful feedback. I especially wish to thank the following individuals, for very useful
comments and suggestions: Louise Antony, Aurélien Darbellay, Heather Logue, Dan
López de Sa, Jennifer Matey, Carl Matheson, Cathy Muller, Bryan Pickel, Sven
Rosenkranz, Pablo Rychter, Elizabeth Schechter, Albert Solé, Chris Tillman, Giuliano
Torrengo, Vera Tripodi, and Bénédicte Veillet.

REFERENCES

Andreasen, R. (1998), ‘A New Perspective on the Race Debate’, British Journal of the
Philosophy of Science, 49: 199–225.
Boghossian, P. (2006), Fear of Knowledge: Against Relativism and Constructivism. Oxford:
Oxford University Press.
Hacking, I. (1999), The Social Construction of What? Cambridge: Harvard University Press.
Haslanger, S. (2003), ‘Social Construction: The “Debunking” Project’, in F. Schmitt (ed.)
Socializing Metaphysics. Lanham: Rowman & Littlefield, 301–25.
—— (2006), ‘Philosophical Analysis and Social Kinds: What Good Are Our Intuitions?’,
Proceedings of the Aristotelian Society, 80: 89–118.
Kitcher, P. (1999), ‘Race, Ethnicity, Biology, Culture’, in L. Harris (ed.) Racism. Amherst:
Humanity Books, 87–117.

© 2013 John Wiley & Sons Ltd

1152 E. Diaz-Leon

Mallon, R. (2007), ‘A Field Guide to Social Construction’, Philosophy Compass, 2: 93–

108.
—— (2008), ‘Naturalistic Approaches to Social Construction’, The Stanford Encyclopedia
of Philosophy. Winter 2008, online at: http://plato.stanford.edu/archives/win2008/
entries/social-construction-naturalistic/
Stein, E. (1992), ‘The Essentials of Constructionism and the Construction of Essentialism’,
in E. Stein (ed.) Forms of Desire: Sexual Orientation and the Social Constructionist
Controversy. London: Routledge, 325–53.

© 2013 John Wiley & Sons Ltd