Social Constructivism and the Philosophy of Science

André Kukla

• Publisher: Routledge
• Year: 2013
• eBook ISBN: 9781134567386
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Social constructionists maintain that we invent the properties of the world rather than discover them. Is reality constructed by our own activity? Do we collectively invent the world rather than discover it? André Kukla presents a comprehensive discussion of the philosophical issues that arise out of this debate, analysing the various strengths and weaknesses of a range of constructivist arguments and arguing that current philosophical objections to constructivism are inconclusive. However, Kukla offers and develops new objections to constructivism, distinguishing between the social causes of scientific beliefs and the view that all ascertainable facts are constructed.

1 Defining constructivism

To say that ‘social construction’ and ‘constructivism’ are vogue words is to understate an extraordinary situation. In a recent survey of the field Hacking (1999) mentions sixty-odd items that have recently been described in print as socially constructed. Here's a haphazardly selected sample: gender (of course), illness, women refugees, quarks, Zulu nationalism, Indian forests, Japan, Ireland, the past, emotions, reality, serial homicide, authorship, the child viewer of television, the Landsat satellite system, dolomite and the self. Hacking notes that the class of putative constructions is not only numerous, but remarkably heterogeneous. Among the items for which constructivist claims have been made, we find people, inanimate objects, states and conditions, events, practices, actions, experiences, relations, substances, concepts and an assortment of what Hacking calls ‘elevator words’ (because they raise the level of discourse, both rhetorically and semantically): reality, truth, facts, knowledge.

What do women refugees, the Landsat satellite system and reality supposedly have in common by virtue of which they're all socially constructed? ‘Don't ask for the meaning,’ Hacking tells us, ‘ask what's the point’ (1999: 5). He describes the point in three clauses:

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 at 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.

(1999:6)

Later on, Hacking adds a zeroth clause:

0 In the present state of affairs, X is taken for granted, X appears to be inevitable.

(1999:12)

In brief, X is asserted to be constructed when we want to call attention to the currently invisible evitability of X for the purpose of mobilizing efforts to evade it.

I can see why conditions (2), (3) and (0) may be said to give the ‘point’ of constructivist claims rather than their ‘meaning’. But what about condition (1)? On even the most traditional accounts of the matter, (1) seems to be a straightforward, necessary condition for the truth of ‘X is constructed’. Hacking himself tells us that condition (1) is something that constructivists about X ‘hold’, as opposed to (2) and (3) which merely specify what they ‘urge’, and (0) which stipulates what they ‘take for granted’. Moreover, Hacking also tells us that conditions (2), (3) and (0) are not absolutely essential prerequisites for constructivist claims. They're merely satisfied ‘very often’. So, despite his disclaimer, Hacking does say something substantial about the meaning of ‘X is constructed’, as opposed to the pragmatic purpose that might be served by asserting it. He says that ‘X is constructed’ entails that X is not inevitable. This may not yet be a full definition; but it is a necessary condition for the validity of constructivist claims.

But is that all that can be said about the truth-conditions for ‘X is constructed’? To say that X is not inevitable is to assert that not-X is possible, and, as every student of philosophy knows, there are various grades and flavours of possibility. Which variety of possibility is at stake here? Hacking is disinclined to pursue this question. His view seems to be that the relevant notion of evitability is not further explicated in constructivist writings: if philosophers want to get clear about what constructivists say without putting words in their mouths, they're going to have to work with the unexplicated notion.

I'm in sympathy with the general Wittgensteinian point that the use of vague concepts doesn't automatically call for remedial treatment. But condition (1) simply doesn't exhaust what can be said about the truth-conditions of ‘X is constructed’, as that phrase is used by those who claim that women refugees, reality and the Landsat satellite system are constructed. Consider: if anything is not inevitable in the vague, unexplicated, everyday sense of the word, it's a freakish accident, like the destruction of New York by a falling asteroid. But nobody would be inclined to say that the freakish accidentality of an event is evidence of its social construction. The possibility of not-X which is relevant to constructivist claims must be delimited in a way that excludes the possible non-occurrence of events of this type.

The needed elaboration of Hacking's condition (1) is given by Nelson (1994). Discussing the special case of the construction of scientific facts, Nelson writes that constructivists endorse the following ‘Constructivist Counterfactual Argument’:

If scientists had chosen to confer facthood otherwise than they actually did, then subsequent history would reflect this in a world-view consistent with the choice they counterfactually made. Therefore, the ‘facts’ are determined by scientists' choices, not by ‘objective reality’.

(Nelson 1994:541)

This formulation is consistent with Hacking's condition (1): the Constructivist Counterfactual Argument entails that the facts of science are not inevitable—they could have been other than what they are. But Nelson further specifies that the evitability is due to the fact that scientists could have made other choices. More generally, the type of possibility at issue in constructivist claims is the option of free agents to do something other than what they actually did. It's true that the social construction of gender differences is predicated on the idea that these differences could have been different from what they are—this much is what Hacking's condition (1) tells us. But that's only part of the constructivist story. In addition, the constructivist thesis about gender entails that gender differences would have been different from what they are if human agents had made different choices. The fact that they could have been different tout court isn't enough. So the whole issue of constructivism versus ‘realism’ arises only in a context where both sides in the dispute accept the good old-fashioned metaphysics of freedom of the will.

In sum, X is said to be constructed if it's produced by intentional human activity. This is more or less what a dictionary would have told us in the first place. It follows that pianos, television sets, cheese sandwiches and all other artifacts qualify as constructed. To be sure, everybody has always known this—Hacking's condition (0) fails to be satisfied by cheese sandwiches. But it's nonetheless a truth, albeit a humble one, that cheese sandwiches are constructed. It should be only slightly less obvious that all our concepts are constructed. The fact that we conceptualize some people as women wouldn't be a fact if we acted differently—e.g., if we didn't conceptualize anybody as a woman. But there's no point making that argument specifically for the concept of a woman. It's just as true for the concept of the colour blue, or for the concept of a quark. The view that concepts are human constructions is sometimes contrasted with the doctrine of natural kinds, according to which only some conceptual schemes manage to carve nature at its pre-existing joints (see for example Hacking 1999:82–4). But the two claims are orthogonal. Suppose that nature has joints. Then one of our concepts may succeed while another fails to carve nature at its joints. But that doesn't make the first concept any less of a construction: if a complex pattern of human activity had been different, we wouldn't have fashioned that particular natural-kinds concept. The fact that it carves nature at its joints is neither here nor there. Trying to carve nature at its joints is just another optional project which we might or might not undertake.

Of course to say that the concept of a woman or a quark is constructed is not yet to say that women themselves, or quarks themselves, are constructed. The claim that these entities themselves are produced by intentional human activity is substantially stronger. In the case of women, it's easy to see how women might turn out to be constructed. Here is one possible (and entirely unoriginal) scenario. We begin by constructing the concept of a woman. We include in this concept all the traditional appurtenances of femininity: nurturance, seductiveness, social intelligence, a poor sense of direction, and so on. Naturally, those to whom this concept is applied come to know that the concept is applied to them. This knowledge leads them to behave in ways that are different from how they would have behaved if they had not been so categorized. Perhaps it causes them to have a poor sense of direction by undermining their self-confidence. The result is the social construction, not just of the concept ‘woman’, but of women. Women turn out to be a type of being that wouldn't exist if a certain pattern of intentional human activity had not taken place.

It's not so easy to devise an equally commonsensical scenario whereby the facts about quarks (not just our concept of a quark) turn out to be socially constructed. But it's precisely claims of this type—more generally, claims that the facts of so-called natural science are constructed—that I'll be dealing with in this book.

Before we start, it's necessary to distinguish three issues that receive a great deal of play in the constructivist literature. People who call themselves constructivists sometimes argue for a metaphysical thesis about some or all facts about the world we live in, sometimes for an epistemological thesis concerning what can be known about the world, and sometimes for a semantic thesis concerning what can be said about the world. The characterization of constructivism immediately preceding this paragraph equates it with the metaphysical claim: women, or quarks, are invented rather than discovered. Most of the chapters in this book are devoted to an examination of various grades of this metaphysical hypothesis.

The epistemological claim associated with constructivism is the thesis of epistemic relativism. This is the view that there is no absolute warrant for any belief—that rational warrant makes sense only relative to a culture, or an individual, or a paradigm. Both the metaphysical thesis and the epistemological thesis are often regarded as two sides of one and the same ‘constructivist’ coin. Thus Fine states that the two doctrines of ‘constructivism’ that give it its philosophical interest are ‘its anti-realism and its relativism’ (Fine 1996:232). And Nelson writes:

philosophical constructivism…is relativistic in two senses. First, there is an ontological relativism about entities and processes. We are not to think of the phenomena studied by scientists as the inevitable manifestations of objectively existing entities and processes; instead, theoretical entities and processes are constituted or constructed by scientists post hoc… The second relativistic facet of constructivism concerns scientific rationality. According to non-relativistic rationalists…defensible scientific decisions, if not correct ones, should be made…in accord with universal standards governing the use of appropriate scientific evidence…Constructivists, holding as they do a sort of relativism about rationality, deny the universality of such standards.

(Nelson 1994:535–6)

My nomenclature is different from Nelson's. I usually reserve the term ‘constructivism’ for what he calls ‘ontological relativism’, and I use ‘epistemic relativism’ to refer to his ‘relativism about rationality’. When there's a danger of being misunderstood, I sometimes refer to Nelson's ‘ontological relativism’ as the thesis of metaphysical constructivism.

Despite their frequently being espoused by the same individuals, (metaphysical) constructivism and epistemic relativism are, at least prima facie, independent doctrines. To begin with, constructivism doesn't obviously entail epistemic relativism. It's (prima facie) possible to combine the constructivist view that facts are socially constructed with the anti-relativist idea that we can nevertheless have absolutely true or absolutely false ideas about them. The fairly uncontroversial thesis that at least some aspects of social reality are constructed provides an apt illustration. The value of money is a socially constructed fact: the pieces of paper that we call money enable us to buy things only because it's widely acknowledged that they enable us to buy things. Nevertheless, an isolated individual who believed that dollar bills have no purchasing power would be absolutely wrong. By the same token, it could be maintained that scientific facts are socially constructed, but that once they've been constructed, it's a mistake for anyone to disbelieve them. The conceptual option of avowing constructivism while denying relativism seems to have been taken up by some constructivist authors. Latour and Woolgar, whose Laboratory Life is one of the most influential documents in the literature of constructivism, caution the reader that their position ‘is not relativist’ (1986:180).

Conversely, it's possible to combine the view that beliefs are only relatively warranted with the anti-constructivist hypothesis that there is an independent reality. This is the position that Devitt (1991) calls ‘fig-leaf realism’. Fig-leaf realists admit only that something exists independently of human activity, but they deny that we can have absolute knowledge of any of its properties. Kant was a fig-leaf realist. So is the contemporary sociologist and relativist Karin Knorr-Cetina (1993:557). Knorr-Cetina still counts herself among the ‘constructivists’, however, partly on the basis of her relativism, and partly on the grounds that she regards the relativized facts of science as constructed.

The second type of claim to be distinguished from the metaphysical thesis is a semantic hypothesis. Constructivists are wont to say that nature ‘plays no role’ in the acceptance of scientific claims. When they say this, they sometimes have in mind the metaphysical thesis: nature plays no role in scientific acceptance because it's prior acceptance that constitutes, as it were, the nature of nature. At other times, however, the argument takes a decidedly semantic turn. Harry Collins (1985), for instance, maintains that past verbal usage doesn't determine the future application of words. It follows that sentences have no determinate empirical content: there is no fact of the matter whether an event confirms or disconfirms a hypothesis. The outcome can be negotiated either way. On this account, nature plays no role in scientific acceptance because it fails to hook up to language in the requisite manner. This semantic constructivism is, at least prima facie, independent of both (epistemic) relativism and (metaphysical) constructivism. The semantic thesis refers to sentences, while the relativist thesis refers to beliefs. It's possible to reconcile semantic constructivism with the denial of relativism via the notion of tacit knowledge: sentences have no determinate empirical content, but we may still have non-propositional knowledge about the world that's absolutely correct. That is to say, we may tacitly know what happens next and act accordingly, even if we're unable, even in principle, to say what happens next. The converse proposition—that relativism doesn't entail semantic constructivism—is intuitively compelling. The fact that warrants for belief are all relative doesn't mean that there aren't any absolutely true sentenc...

Table of contents

• Cover
• Social Constructivism and the Philosophy of Science
• Philosophical issues in science
• Full Title
• Copyright
• Dedication
• Contents
• Preface
• Acknowledgements
• 1 Defining constructivism
• 2 Constructivism and the sociology of scientific knowledge
• 3 The varieties of dependence
• 4 The varieties of constitutive constructivisms
• 5 The empirical case for constructivism
• 6 The a priori case for constructivism
• 7 Three brief and inadequate objections to constructivism
• 8 The problem of misrepresentation
• 9 Constructive empiricism and social constructivism
• 10 The infinite regress of constructions
• 11 The Duhemian asymmetry
• 12 The problem of the two societies
• 13 Constructivism and time
• 14 Constructivism and logic
• 15 Relativism
• 16 Semantic constructivism
• 17 Irrationalism
• 18 Conclusions
• References
• Index