Introduction

Among the many notable aspects of Peter Berger and Thomas Luckmann’s (1966) The Social Construction of Reality is that it is the first book to use the terms “social construction” in its title (Hacking 1999: 24). Their basic argument is that stable social institutions emerge from highly flexible possibilities at the individual and interactional level, which become externalized and objectivated, eventually being taken for granted as realities. New cohorts are socialized into (and through) such institutions, and deviance is managed through social control processes, so that individuals internalize and identify with the institutional norms and roles that circumscribe their actions. Knowledge, in a very broad sense of the term, runs through this circuit from individuals to institutions, and back, though Berger and Luckmann add that individuals typically do not recognize that the institutions they treat as external realities are human constructs, and as such could (and under different conditions do) differ from the specific forms they currently take in a given society. Consequently, for well-socialized members of such a society, institutions that originated through social actions (as constructed orders) are now taken for granted as objective and unchangeable, supported by natural and/or supernatural laws.

In the decades that followed the publication of Berger and Luckmann’s landmark work, countless other academic books and articles adopted “social construction” in their titles, to the point of making up a movement that ran through a broad array of fields in the humanities and social sciences. In this essay, I will discuss a prominent and highly influential line of constructivist work that helped coalesce the interdisciplinary field of “Science and Technology Studies” (STS—the same acronym also stands for the alternative name, “Science, Technology and Society”). Social constructionism in STS was particularly prominent and controversial in the 1980s and 1990s. And, while social constructivism is no longer as prevalent as it once was in book and article titles or conference programs, it left an indelible mark on the field. The aim of this essay is to twofold: first, to clarify the relationship between social construction in STS and Berger and Luckmann’s original conception of it; and, second, to point to persistent problems and confusions that arise when social construction is expanded indiscriminately to encompass all natural as well as social orders.

From Social Construction to Generalized Constructivism

Ian Hacking (1999) points out that the version of social construction that Berger and Luckmann discussed had no direct connection to earlier constructivist movements in art and mathematics, or with systematic methods of theory or proof construction.Footnote 1 It also is the case that the ventures into social constructivism that followed in the wake of Berger and Luckmann’s book often had little to do with the original version, and they often differed from one another. Berger and Luckmann treated “social construction” as a substantive theme, and they did not promote a general philosophical perspective (an “ism”). However, the words “social construction” had an appeal that went well beyond any particular theory or philosophy.

Hacking provides a schematic account that, despite its severe simplicity, provides a pretty good way to begin characterizing the different varieties. He starts by noting that, where Berger and Luckmann place reality in their title, other titles replace that comprehensive term with an amazing variety of specific topics: social problems, mental disorder, illness, knowledge, nature, and Zulu nationalism, among many others. Hacking’s formula uses X to encompass the items on this list. The first, most basic, variant, of the formula is: “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”. A second and third formula takes us further, into normative and politicized versions of constructivism: “X is quite bad as it is”; and “We would be much better off if X were done away with, or at least radically transformed” (Hacking 1999: 6).

As variants of constructivism proliferated, they became increasingly remote from Berger and Luckmann’s theoretical program. Constructivism (true to its own theoretical articulation) took on a life of its own, to the point that when philosopher John Searle wrote a book with the slightly transformed title The Construction of Social Reality (1995) he apparently was unaware of Berger and Luckmann’s contribution.Footnote 2 Viewed with hindsight, Berger and Luckmann’s theory, though very broad in scope, was far more circumspect than many of the explicitly constructivist ventures that followed in its wake. The reality in their title was social-institutional reality, ranging in scope from world religions to the intimate routines of the household dinner table and bedroom. They drew upon a broad range of classic social theorists, including Marx, Durkheim, Weber, Mead, Simmel, and Schutz, and a substantial portion of their book reads as a primer in sociological theory, more advanced than, but not unlike, Berger’s Invitation to Sociology (1963).

Berger and Luckmann’s theory holds critical implications, as it undermines reified and naturalized treatments of social identities and institutionalized activities, and encourages the conviction that such identities and activities need not remain as they are. However, Berger and Luckmann neither suggest that socially constructed institutions are necessarily “bad” nor that such institutions should be reformed or overthrown because they are socially constructed. Given the general scope of their discussion of social reality, it would make no sense to indiscriminately denounce and seek to overthrow such reality. Berger and Luckmann also make clear that socially constructed reality does not encompass all reality. They distinguish between natural and social reality, and while they do not draw strict lines of demarcation, the distinction is crucial for setting up their argument. In their more explicitly philosophical treatments of social construction, Hacking and Searle also use variants of that distinction, and argue that the presumption of an un-constructed counterpoint is necessary for understanding the distinctive ontology of socially constructed phenomena. For them, as well as Berger and Luckmann, socially constructed phenomena, such as the value of currency, the division of labor, and actively enforced legal statutes, are not an illusory form of ‘real reality’ that is merely ephemeral or imaginary and less objective than rocks in the field and physical laws.

Berger and Luckmann devote less attention to ontology, and much more to sociology, as the theoretical problem they take up is how social institutions are established as intersubjective, persistent, and external orders, with an existence that inhabitants of a given society encounter as facts of life. In one sense, these facts are no less constraining than physical forces and biological needs, but in another sense they depend on unreflective compliance with institutionalized patterns of conduct. For Berger and Luckmann, social facts depend for their existence on historically and culturally variable conventions, socialization processes, and concerted human practices. The most provocative aspect of their argument is that elites who benefit from socially constructed institutions encourage the populace to take them for granted as immutable products of natural and/or supernatural forces. As children of the Enlightenment who have access to the theories and findings of empirical sociology, Berger and Luckmann argue that it is only through independent and cosmopolitan reflection that the socially constructed origins of such institutions can be exposed. They devote much of their book to demonstrating that the symbolic systems that every civilized society creates and sustains naturalize and universalize social facts in a way that obscures their constructed origins and sustains the status quo ante. Although they distinguish the socially constructed domain from that of nature, and remain non-committal about supernatural agencies, their humanistic treatise heretically suggests that much of what, in their own society as well as others, is officially designated as immutable natural and supernatural law, is actually a legacy of contingent human action.

Berger and Luckmann’s approach to social reality is humanistic, in the sense that it avoids using explanations of social order that invoke supernatural and other non-human agencies. In line with classical sociological theory, they treat social order as the product of institutionalized actions, occasionally formulated by explicit rules and individually enacted through internalized social roles and norms. Such rules, roles, and norms are both conventional and constraining: they can change and are changed both by official rulings by legislative bodies and in a more gradual and casual way through cultural drift and local negotiation. The enforcement of particular rules can be loose or strict, arbitrary or precise, and open to dispute or not. And, Berger and Luckmann argue, the elites and intellectual authorities in any stable society eventually come to support the rules they hold most dear with legitimations that naturalize, moralize, and deify their authority. So, for example, although sexual codes and norms are highly variable across time and culture, in any given historical society violations of the most fundamental normative orders are denounced and punished as offenses against nature or God.

Berger and Luckmann distinguish their explanatory approach from a nomothetic approach, which they view as appropriate for explaining order in nature but not in society. Using their approach thus requires us to make a judgment about which phenomena are socially constructed and which are not, a judgment that is particularly difficult to make when faced with phenomena that are authoritatively presented in our own society as “natural”. This difficulty is less severe in a secular and pluralistic academic context when it comes to religious doctrines, culturally specific moral stipulations, and political ideologies, but much more challenging when it comes to natural facts and laws authorized by established science. This challenge has become acute in the half-century following the publication of The Social Construction of Reality.

Social Construction Invades Natural Order

A decade after Berger and Luckmann published their book, social constructivism began to take hold in the sociology and social history of the natural sciences. The first significant move in this direction was inaugurated in the new Science Studies Unit at the University of Edinburgh. David Bloor (1976), a member of the Unit, announced a “Strong Programme” in the sociology of knowledge. Bloor’s key proposal was to expand the explanatory scope of the sociology of knowledge to encompass natural order as well as social order. He did not adopt the term “social construction,” or even mention Berger and Luckmann in his prolegomenon, but like them he adopted established explanatory concepts such as interests, conventions, traditions, and socialization processes. Unlike them, however, he suggested that these concepts were appropriate not only for explaining religious doctrines, popular superstitions, and political ideologies, but also for explaining the emergence and current acceptance of established scientific and mathematical knowledge. In other words, he proposed that an explanatory approach along the lines of the one that Berger and Luckmann used for explaining social order was equally appropriate for explaining the origins of even the most robust natural laws and mathematical procedures—those that are accepted today in the natural sciences and used as a basis for further inference and action. Because socially constructed realities are understood to vary with the historical societies that construct them, Bloor’s position required a subscription to a relativistic view of science and mathematics, which became a subject of debate with philosophers of science (Hollis and Lukes 1982).

Some insight into this broadened application of social constructivism can be gained by considering the way Berger and Luckmann explain religious doctrines. Using a fanciful example (they use many such examples), they invite readers to imagine a dispute between two factions of “dervishes” about the ultimate nature of the universe. Initially, the dispute occurs in a remote desert, far from the surrounding society, and is only of interest to the participating adepts. However, if this esoteric dispute catches the attention of more mainstream elements of the society, “it will be largely extratheoretical interests that will decide the outcome of the rivalry. Different social groups will have different affinities with the competing theories and will, subsequently, become ‘carriers’ of the latter” (Berger and Luckmann 1966: 120). Once adopted, the socially accepted theory will attain the status of true knowledge, while the competing theory is consigned to error or heresy. A similar style of explanation was used by Steven Shapin (1975; at the time, a colleague of Bloor’s in the Science Studies Unit at the University of Edinburgh), in a study of an early nineteenth century dispute between phrenologists and anti-phrenologists. Like Berger and Luckmann in their account of the dervish dispute, Shapin is indifferent to the truth or falsity of the competing theories. To maintain such indifference is more difficult in his case, since phrenology is now widely regarded as an example of a pseudoscience, whereas Berger and Luckmann do not even describe the contents of the competing dervish theories. However, like Berger and Luckmann, Shapin does not treat the intrinsic truth or falsity of phrenology as a relevant factor in his explanation, and instead turns to extratheoretical interests to explain the outcome of an extended historical episode in which it was contested and eventually rejected by university authorities. He relates support and opposition to phrenology to different socioeconomic sectors in Edinburgh’s rapidly changing society, and argues that phrenology had particular appeal for members of the ascendant “working and lower middle-classes peculiar to Edinburgh” (1975: 227) in the early decades of the nineteenth century. Using Berger and Luckmann’s language (adopted from Max Weber), one could say that these classes were “carrier groups” with an elective affinity for phrenology; groups whose members found appeal in the phrenologists’ “outsider” status and their emphasis on individual qualities, unlike their opponents who were differentially drawn from more established professions and elements of the old aristocracy, and who emphasized a moral philosophy supportive of tradition. The latter groups remained in control of key positions at the University of Edinburgh, and eventually succeeding in suppressing the ongoing debate within the confines of that university—a crucial move for de-legitimating phrenology. Shapin’s explanation goes into further complications and is much more elaborate than Berger and Luckmann’s vignette, but for present purposes it illustrates how a sociological explanation much like one they use for a religious dispute was applied to a dispute in a “scientific” domain.

A few years later, “social construction” was used in the subtitle of Bruno Latour and Steve Woolgar’s (1979) Laboratory Life: The Social Construction of Scientific Facts. This book was framed—with strong hints of parody—as a classic anthropological narrative, in which an articulate stranger from abroad drops into a tribe, and refrains from preconceptions about native life while taking copious notes about what he witnesses. In this case, the tribe was a laboratory at Salk Institute, the tribal beliefs were communicated at the laboratory bench and inscribed into the scientific literature, and various rituals were conducted with laboratory implements and sacrificial animals. Much of their description went roughly along the lines of Berger and Luckmann’s theoretical account of “symbolic universes” and their associated “conceptual machineries” of myths, rituals, and sacred doctrines. And, most controversially, Latour and Woolgar used the theme of “social construction” to describe the history of practices that went into the identification and characterization of the chemical structure of a growth hormone, for which the 1977 Nobel Prize in Physiology and Medicine was awarded jointly to the lab director (Roger Guillemin) and the director of a rival lab (Andrew Schally). Latour and Woolgar did not suggest that the discovered “fact” was actually a “myth”; instead, they reconstructed the history of the research, focusing on the historical and practical contingencies, the rivalry between Guillemin and Schally, and the emergence of the “fact” as initial qualifications and uncertainties in the way it was linguistically formulated gradually dropped away. Using an argument rhetorically similar to Berger and Luckmann’s, they suggested an alternative way to summarize the history of a scientific fact: rather than testifying to the gradual and contingent emergence of certainty (both within the lab and the scientific field at large) about the intrinsic chemical structure and physiological action of the peptide in question, their history testified to the institutionalization of a “fact” that could, in principle under as yet unknown circumstances, be undone and redone differently in the future. Accordingly, the “discovery” was a social fact, and the characterization of the peptide and its physiological action that was officially granted the status a natural fact ascribed to nature’s agency did not remove it from the domain of socially constructed reality. Accordingly, the forgotten history of social construction bolstered the stability of the fact and protected it from further critical scrutiny, but with no ultimate guarantee.

Latour and Woolgar did not suggest that laboratory members deliberately set out to deceive their readers when they presented their naturalistic accounts and described their procedures as following step-by-step methodological procedures, but they did draw a striking contrast between the “disorder” of daily laboratory life and the practically and rhetorically constructed “order” that eventually was presented in published reports. This contrast became a prime mover for general constructivist arguments that dramatized the difference between rhetorically presented “objective facts,” and the “messy” laboratories and contingencies that were “deleted” from official versions of those “facts”. This contrast set up proposals to restore the contingent and constructed histories of innovation that had been deleted from official scientific reports and journalistic accounts of discoveries and inventions.

The Proliferation of Social Constructionism in STS

Through the 1980s and into the 90s, social constructivism became a central interpretive scheme for the new field of Science and Technology Studies (STS), as well as for many other areas of sociology, anthropology, and communication studies. In the humanities, social constructivism was interwoven with post-structuralism, psychoanalysis, critical theory, and feminist epistemology, to become part of cultural studies of science (Keller 1984; Rouse 1992; Haraway 1991). A recurrent theme that ran through much of this movement was the deconstruction of the distinction between nature and society—the very distinction that originally circumscribed the domain of social construction. No longer exempted from critical social, cultural, and literary scrutiny, the authority of science became no different from other sources of authority, and the contents of science became thoroughly infused with the privileged cultural perspectives of the predominantly white Northern European and North American men who are honored in the annals of the science.

A symbolic gesture that marked the erasure of the distinction between nature and culture was Latour and Woolgar’s removal of the word “Social” from the subtitle of the second edition of their book, now titled: Laboratory Life: The Construction of Scientific Facts (1986). One of their reasons for this erasure was that constructivism now spanned the previously assumed divide between natural and social order. Another reason had to do with Latour and Woolgar’s conviction that “social” explanations were too limited in scope to resolve the “agency” of non-human entities that have a conspicuous role in scientific practice (including both the things studied and the technologies used for investigating them). This small symbolic gesture marked both the popularity that constructivism had attained and the beginning of its demise—or, at least, the demise of any version that had much resemblance to Berger and Luckmann’s sociological theory.

In the history of the social sciences, there have been repeated efforts to explain social phenomena with natural scientific concepts and principles, often drawn from evolutionary biology and the neurosciences. Unlike such efforts to naturalize sociology, the constructivism that emerged from STS and cultural studies was a mirror image: a thoroughgoing enculturation and politicization of nature. Latour’s (1987) Actor-Network Theory is perhaps the most persistent and prominent of these efforts in STS. That theory is, in part, a reaction to efforts to describe scientific developments sociologically, such as in Shapin’s account of phrenology, or in Latour and Woolgar’s earlier ethnography. Those earlier accounts placed the accent of reality on social interests and social actions and remained non-committal about the intrinsic “natural” status of the phenomena question. But, while he rejects social explanation, Latour (1987) does not accept naturalistic scientific claims on their own terms. Instead, he advises his readers to expand the domains of action and agency beyond the field of human actions, so that humans, collective agencies, non-human entities, theoretical causes, laboratory apparatus, and virtually all other nominal ‘things’ placed in subject position in narratives about scientific and technological innovations, become “actors” (or “actants”) that have “agency”—they are neither passive materials subject to human malleability nor constituents of an impassive nature.

A preliminary understanding of this pan-constructivist ontology can be illustrated with two of Latour’s examples: microbes and speed bumps (often called “sleeping policemen” in British vernacular). According to Latour (1986), Louis Pasteur did not simply discover the microbial basis for diverse problems and processes, such as fermentation, the spoilage of milk and the manufacture of cheese, and diseases such as anthrax and rabies. Instead, he domesticated and disciplined the microbes, as they reciprocally resisted and, only after much negotiation, complied with his administration. His communications with them, and they with him, were achieved through practical and instrumental intermediaries. Moreover, after his microbes were identified, named, cultivated, and denatured, they become fully-fledged social actors whose presence and involvement in various social transactions (sexual, commercial, communicative) and infrastructures (water, sewage, transportation) were progressively disclosed, channeled, and disciplined. The microbial actor did not simply pre-exist these involvements as a hidden, as-yet unknown, cause of infectious disease, spoilage, and pollution; instead, it came on the scene as a novel participant in social actions as a result of Pasteur’s cultivation of it, and of his and his publicity agents’ dissemination of the news about this actor’s debut in medicine, hygenics, and commerce.

The speed bump is a simple “traffic calming” mechanism: a constructed bump in the road designed mechanically to deter vehicles from going too fast. In Latour’s (1992) account of this “mundane technology,” the sleeping policeman takes on an assignment that materially embodies the traffic cop’s role in enforcing rules. It stands (or “sleeps”) as a proxy for the traffic cop’s injunction, “Slow down!”. But, rather than operating as a rule and role-playing enforcer of the rule, it acts materially and mechanically to achieve moral order, while also changing the concrete (literally in this case) division of labor in society.

Actor-network theorists stress that they do not aim to reduce their networks to a monistic array of equivalent actors: the actors retain their specific identities, despite the erasure of the “great divide” between nature and human society and the pervasive relationality that runs through everything.

Backlash

Berger and Luckmann did not set out to debunk the socially constructed doctrines and institutions they described and imagined, but their emphasis on “extratheoretical” human interests and secular functions were incommensurable with the particular supernatural doctrines and political ideologies they characterized as constructions. This incommensurability with the systems of knowledge explained had long been a source of contention with sociology of knowledge explanations. Similarly, when sociologists of scientific knowledge expanded the domain of social constructivism to encompass currently accepted scientific theories and facts, their explanations conflicted with unqualified accounts of nature’s agency. In the 1990s, a number of natural scientists and mathematicians, supported by many philosophers and social scientists, publicly took issue with social constructivism, as well as with feminist and cultural studies of science, in an episode that some journalists dubbed with the hyperbolic label “the science wars” (Gross et al. 1996).

Although proponents of social and cultural studies of science denied that they were promoting an anti-scientific doctrine in league with creationism or climate skepticism, their protestations often were dismissed and ignored. One source of the confusion that enveloped constructivism was the very term “construction”. Disputes about empirical findings often turn on questions about possible artifacts arising from the instrumental and graphic display and analysis of data (Lynch 1985). Unlike deliberate efforts to manufacture evidence in cases of fraud, such inadvertent constructions are usually treated as excusable and even inevitable. Research artifacts range in severity from noise and cosmetic defects that detract from the clarity of evidence, and intrusive features that obscure the visibility of phenomena, to systematic sources of distortion and illusory appearance that can be mistaken for actual entities. So, for example, if recently reported observational evidence for gravity waves were to turn out to be an “artifact,” this would mean that the measurements that had been treated as evidence of minute fluctuations in space–time caused by a massive collision between black holes in deep space, would now be consigned to uncontrolled noise in the highly delicate instrumentation.

For scientists who associate talk of “construction” with skepticism about findings, indiscriminate uses of the term in reference to scientific activities and evidentiary products would seem to express an attitude of global disbelief with an assertion that all scientific measurements are artifacts. And, when constructivism is intensified with vocabularies drawn from normative and critical accounts of social exploitation, inequality, domination, and discrimination, the connotations of skepticism can seem to cross over into denunciations for projecting politics, as well as sexism, racism, and homophobia, into remote, non-human domains.Footnote 3

Dissolution

Although there were some efforts to cut through the confusion and hostility associated with the “science wars” (e.g., Labinger and Collins 2001), the dispute eventually played itself out with no clear resolution. By that time, STS was well established as an academic field, so that, while damage was done to some individuals and programs, the field as a whole was not deeply threatened by public denunciations of the more provocative claims made by its leading proponents. By then, constructivism already was falling out of vogue, as much of the research in the field had become concerned, not with giving social constructivist accounts of well-established laws of physics, but with describing and taking part in current controversies about genetics and biomedicine, the pharmaceutical industry, and potential environmental catastrophes, all of which are closely bound up with contentious social and cultural matters in which uncertainties, fraught matters of policy, and charges of racial and gender bias are close to the surface. Activist STS is in the ascendency as constructivist STS fades.

The legacy of social constructionism still has some hold, however, as STS analyses continue to deconstruct objectivist and naturalistic claims made by “establishment science”. However, in cases such as climate change or creationism, partisan arguments with a family resemblance with those of constructionists are themselves used to prosecute controversies in the face of “orthodox” scientific consensus (Lynch 1998; Latour 2004), raising dilemmas about how to articulate a conception of science congruent with the arguments and evidence an advocate aims to support (Lynch and Cole 2005). The long-standing effort in the sociology of knowledge to formulate a general, non-evaluative theory of knowledge, has once again failed to transcend the uses of social explanation to debunk particular knowledge claims.

Conclusion

Berger and Luckmann avoided the confusions and dilemmas associated with expanding social constructionism into currently accepted, scientifically authorized domains of natural knowledge. But, while they limited their theory to socially constructed phenomena, they pressed it into contentious territory by suggesting that some currently objectivated doctrines and institutions could, and should, be construed as contingent social constructions. By doing so, they set up an implicit credibility contest between their own sociological theory (which was mainly a synthesis of classic sociological theories) and the doctrines they explained by means of it. For the most part, their book presented real or imagined cases of doctrines that secular and cosmopolitan readers (or at least many of them) would agree were historically contingent legitimations supporting powerful institutions and their beneficiaries.

The rapid expansion of constructivism to encroach upon contemporary scientific knowledge placed its proponents in a contest with disciplines that have secured far more credibility than sociology currently holds in contemporary academic (and other secular) institutions. Fortunately, such institutions (or at least some of them) cultivate a reputation for liberalism. While disputes about constructivism did break out in the 1990s, they did not deter the international spread of social and cultural studies of science through substantial areas of the humanities and social sciences. The popularity of social constructivism, as a particular theoretical perspective, waned not because it was heretical, but because, in the sectors of the academy where it was tolerated, it was not heretical enough to suit proponents of a more thorough and relentless politicization of scientific knowledge.