Review of Wiebe E. Bijker, Of Bicycles, Bakelites, and Bulbs: Toward a Theory of Sociotechnical Change, MIT Press, Cambridge MA, 1995.
In many ways the sociology of technology is a sociology of obduracy. Much of the literature in this diverse field asks in one way or another, ‘Just how is it that artifacts reach and maintain their shape?’ Others ask a similar question, but - preferring to engage in a politics of obduracy - ask why such arrangements have occurred. In his new book, Of Bicycles, Bakelites, and Bulbs, Wiebe Bijker offers us a thesis that sits most comfortably in the latter camp. A detailed constructivist history of three important technologies, this book beautifully reviews and sets out the framework of the approach known as the Social Construction of Technology (SCOT). It gives us, as Bijker would argue, a political constructivist theory of technology.
Split into five chapters - an introduction, a conclusion, and three empirical chapters between - the book brings together the various threads of Bijker’s programme over the last decade. The empirical chapters are enlarged and heavily reworked revisions of three earlier papers on bicycles, celluloid and Bakelite, and high intensity fluorescent lighting respectively (Pinch & Bijker 1984, Bijker 1987, Bijker 1992) - all important in the so-called new sociology of technology. The conclusion revises a later paper in which Bijker calls upon scholars of science and technology studies to revisit the political relevance of their early endeavours (Bijker 1993). The first empirical chapter stems from the most influential of the earlier papers - Pinch & Bijker’s (1984) account of the social construction of the bicycle. In the book he sets out in much more detail the relevant social groups (RSGs) that were involved in fixing the dominant meaning of the bicycle. In the example of the stabilization of the Ordinary (or Penny Farthing), he shows how one of the RSGs - young adventurous men - saw the Ordinary as a sporting machine “that was rather hazardous to ride”. (p47) That the Ordinary was seen as risky to ride was however one of its attractive features, so that:
Young and often upper-class men could display their athletic skills and daring by showing off in the London parks. To impress the riders’ lady friends, the risky nature of the Ordinary was essential. Thus the meanings attributed to the machine by the group of Ordinary users made it a Macho Bicycle. (p75)
For another RSG, one that Bijker calls the ‘Ordinary nonusers’, this riskiness was an important reason to avoid riding the perilous machine: “… the machine was difficult to mount, risky to ride, and not easy to dismount. It was, in short, an Unsafe Bicycle.” (p74) Bijker thus deconstructs the Ordinary into two different artifacts: one with the meaning attributed to it by the young daring riders - a Macho Bicycle; the other, an Unsafe Bicycle:
This Macho Bicycle was … radically different from the Unsafe Bicycle - it was designed to meet different criteria; it was sold, bought, and used for different purposes; it was evaluated to different standards, it was considered a machine that worked whereas the Unsafe Bicycle was a nonworking machine. (p75)
Bijker shows here how the ‘working’ and ‘nonworking’ of an artifact are socially constructed rather than intrinsic properties of the object itself. This demonstrates one of the most important notions in the new sociology of technology - that of the interpretative flexibility of artifacts. In other words, in Bijker’s argument, there are as many artifacts as there are RSGs, and there is no artifact not constituted by a RSG. That we can demonstrate the interpretative flexibility of an artifact in this way allows for a sociological explanation of the development of artifacts. If no interpretative flexibility could be demonstrated
all properties of an artifact could be argued to be immanent after all. Then there would be no social dimension to design: only application and diffusion - or context, for short - would form the social dimension of technical development. (p76)
Once the delicate work of deconstruction and interpretative flexibility has been shown, argues Bijker, what then remains to be shown is the closure and stabilization processes of particular artifacts - whether particular technologies become dominant or are lost for ever in a technological scrapyard.
Closure and stabilization are then key concepts in Bijker’s theoretical framework. Aimed at answering the continuity-change requirement, they too first came to light in Pinch and Bijker’s (1984) paper on bicycles. However, Bijker now endows the terms, formerly used interchangeably, with different meanings. Closure is what is in progress when the interpretative flexibility of an artifact is diminishing. It can lead to convergence on a single dominant meaning. It is a social-interactionist process that goes on between RSGs. It is moreover virtually irreversible. In the case of bicycles, closure occurred around the Unsafe and not the Macho interpretation of the Ordinary. Thus the Ordinary became extinct and has next to no chance of being revived. Stabilization, on the other hand, goes on primarily within RSGs and is semiotic, as Bijker would have it. It can be detected when the names used to refer to an artifact become in time simplified. Thus an exhibition report from 1888 refers to “rear-driven safety bicycles” which can have diamond or cross frames. Stabilization in the ensuing years meant that by 1889 “safety bicycle” was enough and by 1895 “bicycle” all by itself would do to label essentially the same artifact, a low-wheeled machine with rear-driving chains and a diamond frame.
The technological frame (TF) appears to be the theoretical improvisation of which Bijker is most proud and its exposition comes appropriately in the longest chapter, which looks at the “invention” by Leo Baekeland of Bakelite. A TF is an arena in which the meaning attributed to an artifact by the members of a RSG is negotiated. It is where the problems of the artifact are identified, and where solutions are proposed and essayed through problem-solving strategies characteristic of that frame. It is the site also of goals, technological and scientific theories, tacit knowledge, testing procedures, as well as design methods and criteria. Is it a catch-all concept? No, says Bijker, because “it is possible to give a quite unambiguous characterization of the technological frame for a relevant social group” (p124). And “describing the technological frame of a relevant social group does indeed explain a particular course of events”.
The TF is to sociotechnical change what Kuhn’s paradigm is to scientific revolutions, though it differs in important respects. For one, it is not narrowly cognitive but materially and socially heterogeneous. Also it is a concept to be applied not only to scientists and engineers but to all RSGs. With the TF the analyst is able, according to Bijker, to breathe life into the sequence of snapshots that alone is possible with only the theoretical concepts of the RSG, interpretative flexibility, closure and stabilization. So it is much more than a symptom, or indeed cause, of artifactual stability. For Bijker it will be helpful in transcending the distinction between hitherto irreconcilable opposites: the social shaping of technology and the technological impact on society, social determinism and technical determinism, society and technology (p196-7).
Once he’s entered it into the drama, Bijker feels able to claim that he has met the four requirements for his theory of sociotechnical change that he laid out in his introductory chapter. Along with closure and stabilization as well as inclusion, it accommodates both technical change and continuity. It combines elements of agency and structure. It transcends the need to use a priori distinctions between the social and technical. And, finally, it allows for the question of whether an artifact “works” to be treated as explanandum rather than explanans.
Technological frames abound in Bijker’s Bakelite history. One develops around celluloid which prominently includes practices and problems associated with the solvents used in its manufacturing process. Another has synthetic-dye chemistry as its focus. The purification and chemical analysis of reaction products is central to this TF’s modus operandi. Other TFs, in which Baekeland becomes strongly included before his Bakelite days, are those of photochemistry and electrochemistry. Early in his career Baekeland, a keen amateur photographer in the early days of photography, worked on innovations in photographic paper and its development. So successful was he that he sold his photographic paper affairs (and an undertaking to stay out of the photographic business!) to Eastman Kodak in 1899 for the then enormous price of $750,000. He became involved in the then burgeoning electro-chemistry TF later, when he worked as a consultant to a firm intent on producing caustic soda from brine. Both frames notably favoured meticulousness and systematic investigation.
His low degree of inclusion in the TF of the synthetic-dye chemists and his high degree of inclusion in the electrochemistry TF meant that Baekeland worked unusually when he eventually turned to experimenting with the phenol-formaldehyde reaction. It had already attracted synthetic-dye chemists but they had dismissed it because its products proved impossible to purify and analyse. Baekeland however paid little attention to analysis. Instead he went about varying the reaction variables systematically. He managed thus to produce several promising chemicals. Some were soluble and interested him and his assistant as potential varnishes; others were insoluble and came to interest him as potenial moldable materials. It was some of the latter which later he and the interactions of multiple RSGs through their TFs stabilized as “Bakelite”.
It is ironic that Baekeland’s retrospective account of his “invention” portrays it in a very different way, as a more or less single-minded journey towards the goal of creating a moldable plastic. Bijker’s tells us why. Refusing psychologism, his is an explanation once more in terms of the versatile TFs. By the time Baekeland came to write his story, long after the “invention”, his inclusion in various TFs had changed:
His degree of inclusion in the celluloid technological frame increased, and the history of the invention of Bakelite was cast in terms that were in accordance with that frame: building on Kleeberg’s work [which had demonstrated an insoluble solid product of the phenol-formaldehyde reaction], aiming at a molding plastic, focus [sic] on the solvent (p150).
By the time we get to the fluorescent-lighting case study, we have seen all of the theoretical concepts that make up Bijker’s analytical framework. So at last we see his approach in its fullness as he explains why even today we use high-intensity and not high-efficiency fluorescent lamps. It is a story that tells of several relevant social groups. On the production side there were the Mazda companies, which dominated the market for electric lighting; the electricity utilities, which feared that high-efficiency fluorescent lamps would slash electricity consumption; the independent lighting manufacturers, burdened with prohibitive patent licences from the Mazda companies; and, later in the story, lighting-fixture manufacturers producing standard fixtures for Mazda lamps. Last but not least of the RSGs was the ‘general public’, which included a subgroup that bulked large in the calculations of some technologists - women. According to a 1940 article in Electrical Engineering green and yellow “mercury lines in the spectrum of fluorescent lamps” meant that “overemphasis of these colors in the wall finishes, the draperies, upholstery, and the complexion often results in unattractive effects very disturbing to the lady of the house” (p246).
We use those high-intensity lamps for a host of reasons, naturally. But one of the more salient ones is, for Bijker, the Nela Park Conference. Here representatives of the Mazda companies and the electricity utilities met to sort out their differences. They did this in an extraordinary way. They designed the now familiar lamp around the conference table. It was a design not yet built and not one whose feasibilty they could be certain of. But the new design satisfied their various interests. It went some way to settling the ‘load issue’, which had the utilities fearing that electricity consumption would plummet because of the high efficiency of recently developed fluorescent lamps. The increased efficiency was compensated for by a much greater brightness (and electricity consumption). It allowed for power-factor correction in the lamp fixtures. This had been another worry for the utilities, who did not want to provide extra generating capacity unnecessarily. And for the Mazda companies it meant they could get on with selling fluorescent lamps without continually worrying about how the utilities might react to their every move. This “amalgamation of vested interests” was a doubly conventional resolution, lodged as it was within both TFs.
Before he turns to his concluding chapter, Bijker explains how power fits into his framework. He takes as his starting point Giddens’s definition, “the transformative capacity to harness the agency of others to comply with one’s ends” (p262), which avoids the pitfall of seeing power as “stuff”. There are two sides to Bijker’s power coin. Semiotic power is a fixedness of meanings, which is apparent in the ways facts, artifacts, agents, practices and relations are articulated in TFs. It is the “apparent order of taken-for-granted categories of existence” (p263). Micropolitics of power, the other side of the coin, describes “how a variety of practices transforms and structures the actions of actors” (p263), and results in a specific semiotic structure, which, naturally, acts back on the micropolitics. The two aspects are related to closure and stabilization. Closure, the reduction of interpretative flexibility, comes about as an outcome of heterogeneous micropolitical actions to fix meanings and is a first step in making semiotic power. The remaining steps are made by the continuing stabilization process. In closure, the TF as a cage confining its members, exerts semiotic power; as a resource (problem-solving strategies, etc.) for its members, it lends micropolitical power to them.
Actually, power does not so much fit into the theoretical framework as overlay it. Semiotic power and the micropolitics of power do not work alongside the theoretical entities already introduced - the RSGs, interpretative flexibility, closure, stabilization, the TFs and inclusion. “Rather, a description in terms of power strategies functions as a neat summary of processes that were otherwise described in terms of “those entities (p266).
In an autobiographical vein early in the book, Bijker tells how he came to the academic study of technology and society. For him it was a detour. He had been in the science-technology-society movement popular among engineering students in the Netherlands in the seventies. They managed to introduce into school and university curricula critical ways of looking at the relationships between science, technology and society. Then they started to look for empirical and theoretical ‘ammunition’ to help in their struggles with technoscientific authority. Bijker thought he might find some in academia.
He may have been there ever since but he has not forgotten his old political motivations. He takes up the political strand again in the final chapter. For him, though, the obvious route into politics is not the one that leads to science and technology policy studies. That would be to ally oneself with the interests of a specific social group. Instead he advocates a “politics of technology”. It should be “emancipationist rather than instrumental”; it should “politicize technological choices rather than pacify them”, and “problematize rather than absolve” (p280). Bijker, it appears, is as radical as ever.
“Of Bicycles, Bakelites, and Bulbs” is a river fed by two tributaries, Wiebe Bijker’s theory of sociotechnical change and the detailed historical case study. It alternately runs in twisting gullies - impatiently following unexpected courses - and coasts along in grand deep flows. It could serve as an excellent route for a first foray into the territory of science and technology studies. It will also repay renavigation by experienced hands, particularly in its fine reworking of the histories of the bicycle, Bakelite and the fluorescent lamp.
Perhaps it should be said however that while the book collects between two covers the theoretical framework adumbrated in his earlier papers, it does not greatly elaborate that framework. True, closure and stabilization are now distinguished from one another; the relationships between technological frame, relevant social group and artifact are fleshed out a little; and we get the mapping of semiotic and micropolitical power onto the framework. Bijker has lubricated his theory. But there is little else new here.
Moreover, Bijker deflects only some of the criticisms that can be made of this theory. In particular there is the argument that the RSGs used by an analyst are themselves the constructions of that analyst. Drawing boundaries around certain groups is always problematic. Given Bijker’s political intentions it is perhaps not surprising that he barely reflects on this. As we have seen he attempts to empirically justify the use of the RSG by adopting as analyst the categories used by the actors themselves. The world as it exists for the RSG “is a good place for the analyst to begin his or her research. Thus the analyst would be content to use “cyclists” as a relevant social group, but introduce separate “bicyclists” and “tricyclists” only when the actors themselves do so (p48). While we accept that Bijker provides such a strategy as a heuristic rather than in the form of some hard-and-fast rule - and we concur with him in his method - we feel it is valuable not to lose sight of the questions that can be raised while studying the fuzziness of boundaries between groups.
Bijker does this himself in many ways. Indeed central to his work is his explicit aim to fracture the traditional diffusionist image of technical innovation, with its black-boxed technologies passed linearly from engineer to user. He argues, “I have taken an additional step beyond this interpretation by suggesting that it is not only engineers … but all relevant social groups who contribute to the social construction of technology” (p273). Recall his interest in the Nela Park Conference. “At that meeting a third fluorescent lamp was designed - not on the drawing board or the laboratory bench, but at the conference table” (p238). In one move he rubs out the distinction between those who produce and those who use technology - casting the managers as designers. In a futher move he is happy to retain other distinctions, to maintain other boundaries - General Electric as a RSG with specific interests - and with it, the stability of the RSG. Bijker seems to shift between purity and heterogeneity to further his story.
Throughout his case studies he leaves vague his distinction between the social and the technical, but in the conclusion argues for other differentiations which are less reductionist. He argues, “Instead of technical artifacts, our unit of analysis is now the ‘sociotechnical ensemble’” (p274). Every time “artifact” is written, it is shorthand for “sociotechnical ensemble” and we should be able to trace the “[social] construction” of the artifact. Likewise, each time we write “social institution” we really mean “sociotechnical ensemble”, and we should be able to trace the “technical relations” that go into stabilizing that institution. Introducing such complexity (or irreduction) is seemingly at odds with Bijker’s political programme, but he argues for a need to produce “other differentiations between explanandum and explanans, between dependent and independent variable, between foreground and background - differentiations that are not based on the distinction between the social and the technical” (p274). Thus Bijker is arguing for other types of reduction (that which is not social and technical), of the sort where we stabilize enough of this sociotechnical ensemble, so that we can explain the Other against this background. Rather than talk of the social or technical we can talk of the particular configurations of TFs. If we know of the particular configuration (one in which there is for example a single dominant RSG that insists upon its definition of the artifact’s problems and solutions) we are then able to specify what process of sociotechnical change is likely to occur.
In telling his story, Bijker is willing to iron some new creases into the seamless web while ironing others out. For us what is interesting is why he is content to make firm the boundaries around the interests of RSGs such as General Electric - and at the same time unpick the boundaries around others, such as those around design engineers on the one hand and the managers at the Nela Park Conference.
At the very end of the book, Bijker leaves us with an unsettling double paradox. The first he acknowledges directly. Existence is a seamless web in which no individual or collective sociotechnical ensemble is privileged. What then can a theory of sociotechnical change do, he asks, “to help establish institutional and structural ways of guaranteeing the democratic nature of technological culture”? The answer he invites us to accept is to recognize that “there is [and, we presume, can be] no Leviathan” and simultaneously to propose “some form of institutional regulatory system”, even if that is paradoxical. The form of such a system is left entirely to the readers’ imagination. Before the pages ran out as we progressed through the concluding chapter - with its plea for political relevance in academic studies of technology and its advocacy of a politics of technology - we had begun to hope for a more explicit political programme. We were disappointed.
A very mild dose of reflexivity reveals the other paradox. Bijker tells us that his theory of sociotechnical change and stability is tantamount to neutral:
One might attempt to argue that the sorts of STS studies I have discussed, by highlighting the constructed nature of facts, artifacts, social orders, and sociotechnical ensembles, will allow those who are kept hostage by the semiotic power structrues involved … to sever these bonds and free themselves…. Although this may happen, there is no guarantee that it will always work out this way. First science and technology may also be fruitfully employed by the less privileged…. And second there is no reason why the powerful may not draw on the insights of the STS community (p289).
The social construction of technology in that case seems to mirror the technology so often pushed by technologists: something that can be used equally for good or ill. So it is the exception among sociotechnical ensembles. Unlike the technology told of by Bijker, it has no politics.
References
Pinch, T. J. and Bijker, W. E. 1984. “The Social Construction of Facts and Artefacts: or How the Sociology of Science and the Sociology of Technology might Benefit Each Other”, Social Studies of Science 14, 399-441.
Bijker, W. E. 1987. “The Social Construction of Bakelite: Toward a Theory of Invention” in Bijker, W. E., T.P Hughes, and T.J, Pinch (eds.), The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology, Cambridge, MA: MIT Press, 159-187.
Bijker, W. E. 1992. “The Social Construction of Fluorescent Lighting, Or How an Artifact was Invented in its Diffusion Stage” in Bijker, W. E. and J. Law (eds), Shaping Technology/Building Society: Studies in Sociotechnical Change, Cambridge, MA: MIT Press, 75-102.
Bijker, W. E. 1993. “Do Not Despair: There Is Life After Constructivism”, Science, Technology, and Human Values, Vol. 18, No. 1, 113-138
Authors’ addresses: c.stokes@lancaster.ac.uk n.pollock@lancaster.ac.uk