The core idea seems clear enough. To say of something that it is socially constructed is
to emphasize its dependence on contingent aspects of our social selves. It is to say: This
thing could not have existed had we not built it; and we need not have built it at all, at
least not in its present form. Had we been a different kind of society, had we had
different needs, values, or interests, we might well have built a different kind of thing, or
built this one differently. The inevitable contrast is with a naturally existing object,
something that exists independently of us and which we did not have a hand in shaping.
There are certainly many things, and facts about them, that are socially constructed in the
sense specified by this core idea: money, citizenship and newspapers, for example. None
of these things could have existed without society; and each of them could have been
constructed differently had we so chosen.
As Ian Hacking rightly observes, however, in his recent monograph, The Social
Construction of What? (1999), social construction talk is often applied not only to
worldly items – things, kinds and facts – but to our beliefs about them. Consider Helene
Moussa’s The Social Construction of Women Refugees (1992). Clearly, the intent is not
to insist on the obvious fact that certain women come to be refugees as a consequence of
social events. Rather, the idea is to expose the way in which a particular belief has been
shaped by social forces: the belief that there is a particular kind of person – the woman
refugee – deserving of being singled out for special attention.
Talk of the social construction of belief, however, requires some elaboration of the core
idea. For it is simply trivially true of any belief that we have that it is not necessary that
we should have had it and that we might not have had it had we been different from the
way we actually are. Consider our belief that dinosaurs once roamed the earth. It is
obviously not inevitable that we should have come to this belief. We might never have
considered the question. Having considered it, we might have arrived at a different
conclusion, for a variety of causes: we might not have been interested in the truth; we
might not have been as intelligent at figuring it out; we might never have stumbled across
the relevant evidence (the fossil record).
These observations supply various boring senses in which any belief might be considered
dependent on contingent facts about us. The important question concerns the role of the
social once all of these factors have been taken into account: that is, keeping our skills
and intelligence fixed, and given our interest in the question and our desire to learn the
truth about it, and given our exposure to the relevant evidence, do we still need to invoke
contingent social values to explain why we believe that there were dinosaurs? If the
answer is ‘Yes’ – if it’s true that another society, differing from us only in their social 2
values, would have arrived at a different and incompatible belief – then we could say that
our belief in dinosaurs is socially constructed.
It is crucial, therefore, to distinguish between a constructionist claim that’s directed at
things and facts, on the one hand, and one that’s directed at beliefs on the other, for they
are distinct sorts of claim and require distinct forms of vindication. The first amounts to
the metaphysical claim that something is real but of our own creation; the second to the
epistemic claim that the correct explanation for why we have some particular belief has to
do with the role that that belief plays in our social lives, and not exclusively with the
evidence adduced in its favor. Each type of claim is interesting in its own way.
If a thing were shown to be socially constructed in the first sense, it would follow that it
would contravene no law of nature to try to get rid of it (which is not the same as saying
that it would be easy to do so – consider Manhattan). If a belief of ours were shown to be
socially constructed in the second sense, it would follow that we could abandon it
without fear of irrationality: if we have the belief not because there is adequate evidence
in its favor but because having it subserves some contingent social purpose, then if we
happen not to share the social purpose it subserves we ought to be free to reject it.
Much important work has been done under each of these headings, most significantly, it
seems to me, for the topics of gender and race. Simone de Beauvoir (The Second Sex,
1953) and other feminist scholars since, have illuminated the extent to which gender roles
are not inevitable but are rather the product of social forces. Anthony Appiah (Color
Conscious: The Political Morality of Race, 1996, with Amy Gutman) has been
particularly forceful in demonstrating that nothing physical or biological corresponds to
the racial categories that play a pervasive role in our social lives, that these categories
owe their existence more to their social function than they do to the scientific evidence.
Other claims are more controversial. Mary Boyle has argued that our belief in
schizophrenia is socially constructed (Schizophrenia: A Scientific Delusion? 1990). Her
claim is that there is no adequate reason to believe that the symptoms commonly lumped
under this label are manifestations of a single underlying disease and, hence, that the
search for its etiology by neurochemistry is doomed. Perhaps she is right: our
understanding of mental illness is certainly in its infancy. On the other hand, there
appears to be increasing evidence that the symptoms associated with schizophrenia are
predictable significantly before their onset and that the condition is highly heritable.
These facts point in the opposite direction.
In a flourishing research program, we find the expected mix of important and debatable
work. What bears emphasis, however, is that while some particular social construction
claims may be empirically controversial, the templates of which they are instances are in
no way philosophically controversial. Both the abstract thought that some things are
created by societies and the thought that some beliefs owe more to social values than they
do to the evidence in their favor, are as old as reason itself. Whence, then, the
widespread impression that social constructionists are anti-rationalist, anti-realist and
anti-objectivist? 3
The answer is that it stems not from the forms of the claims themselves, and not from
their application to this or that empirically debatable subject matter. It stems, rather,
from the desire of some prominent theorists in this tradition to extend social construction
talk to absolutely everything and, in particular, to the facts studied by, and the knowledge
claims emanating from, the natural sciences. If we are to find our way through the
muddy battleground on which these now famous science wars are being waged it will
help to observe certain distinctions. I will begin with the claim about facts and things.