[Draft. Forthcoming in the Cambridge Journal of Economics]


Marxism and the Critique of Social Rationality:

From Surplus Value to the Politics of Technology


Andrew Feenberg


Abstract: The most effective way to silence criticism is a justification on the very terms of the likely critique. When an action is rationally justified, how can reason deny its legitimacy? This paper concerns critical strategies that have been employed for addressing the resistance of rationality to rational critique especially with respect to technology. Foucault addressed this problem in his theory of power/knowledge. This paper explores Marx’s anticipation of that approach in his critique of the “social rationality” of the market and technology. Marx got around the silencing effect of social rationality with something very much like the concept of underdetermination in his discussion of the length of the working day. There are hints of a critique of technology in his writings as well. In the 1960s and ‘70s, neo-Marxists and post-structuralists demanded radical changes in the technological rationality of advanced societies. Soon technical controversies spread, primarily through the influence of the environmental movement. The concept of underdetermination was finally formulated clearly in contemporary science and technology studies, but without explicit political purpose. Nevertheless, this revision of the academic understanding of technology contributes to weakening technocratic rationales for public policy. A new era of technical politics has begun.



The trajectory to which my title refers suggests my intention of linking Marx’s writings to current technical issues. I am primarily interested in the methodological implications of Marx’s contribution rather than the details of his thought. Please keep this qualification in mind as I discuss the theories of surplus value and deskilling. I will also engage contemporary constructivist technology studies along the way. There are interesting similarities between aspects of Marx’s method and constructivism.

Why Marx? The question is reasonable in the current intellectual context in which Marx’s economic thought is generally dismissed as outdated and his historical significance diminished by the fall of the Soviet Union. But Marx’s engagement with technology has had a continuing influence, in fact, several different influences that are not about to go away.

Technology studies emerged not out of Marx but out of an earlier development of anti-positivist science studies. Epistemological relativism triumphed in the wake of Kuhn’s famous book on scientific revolutions as science came to be studied as a fundamentally social phenomenon.  The rejection of positivism in science studies prepared the way for a break with determinism in technology studies. Technology too, it turned out, was a social phenomenon and not an instrumental application of universal knowledge. Constructivism aimed to liberate us from the last myth, the myth of pure rationality, with the unstated purpose of combating the authoritarianism of the experts.

Constructivism had the good fortune to emerge at a moment of growing concern with environmental issues and the rise of the Internet. As public understanding of technology grew more sophisticated, scholarly study of technology found an unusually wide audience. However, the revival of religious fundamentalism called forth a defense of traditional rationalism. The Enlightenment struggle against superstition and intolerance is not over and the weapons of Enlightenment are still needed, among them the unwavering appeal to scientific truth. The scholarly debate headed toward deadlock as the political context of discussion contaminated the intellectual space with a toxic dose of reality.

I do not intend to address the philosophical issue of relativism versus rationalism in this paper. I do not believe it is necessary to resolve such vast questions to develop a powerful critique of rationality as it is realized in social institutions such as markets and technologies. I will argue that some aspects of Marx’s method point the way to an alternative to the epistemological emphasis of current debates. Widening the context of discussion to include Marx may help us free technology studies from the heavy philosophical burden it inherits from its origins in science studies. Marx allows us to maintain a critical stance in the framework of a social concept of rationality.

I would like to begin by contrasting Marx with Foucault. Foucault is probably Marx’s most influential critic on the left today. His theory is explicitly presented as an alternative to Marxism. Marxism, he argues, still conceives of power as “sovereign,” that is to say, as repressive. The Enlightenment notion that truth is above politics goes along with this. These are, Foucault claims, outdated conceptions of both power and knowledge.

In modern societies knowledge is conjoined with power and together they are productive of individual subjectivity and the social order. Disciplines such as criminology and psychiatry arise along with the institutions of confinement that place their human objects at their disposal. They reshape these objects through disciplinary procedures and so create a modern society.

According to Foucault, power/knowledge is a web of social forces and tensions in which everyone is caught as both subject and object. This web is constructed around techniques, some of them materialized in architecture or other devices, others embodied in standardized behaviors. These do not so much coerce and suppress the individuals as guide them toward the most productive use of their bodies. On this account, technology is just one among many similar mechanisms of social control, all based on pretensions to neutral knowledge, all having asymmetrical effects on social power.

This explains why the social imperatives of modernity are experienced as technical constraints rather than as political coercion. Surveillance, disciplinary power, normalization, all make modern life possible. They "condense" technical and social functions at the level of everyday behavior, even before that functional duality is transferred to the design of institutions and devices. Eventually these constraints are embodied in structures that determine individual’s action more effectively than rules and commands by determining their reflexes, skills and attitudes. The Panopticon is Foucault’s one developed example of the place of artifacts in his theory. Of it he writes,

The exercise of power is not added on from the outside, like a rigid, heavy constraint, to the functions it invests, but is so subtly present in them as to increase their efficiency by itself increasing its own points of contact. The panoptic mechanism is not simply a hinge, a point of exchange between a mechanism of power and a function; it is a way of making power relations function in a function, and of making a function function through those power relations (Foucault, 1977: 206-207).

Foucault claimed that a power based on knowledge and embedded in social techniques and technology cannot be overthrown by a political revolution. The state in its modern form exists through the effects of power/knowledge and a change of policies and personnel would leave these effects intact. What is possible is both less and more than such a change, however revolutionary. Foucault argued for “the subversive recodification of power relations” underlying the structure and methods of the sciences in order to integrate the subjugated knowledge possessed by those on the bottom of the hierarchy (Foucault: 1980: 123). The aim is not to abolish power but to find a way "which would allow these games of power to be played with a minimum of domination" (Foucault, 1988: 18; Feenberg, 2002: chap. 3).

Foucault’s theory is enormously suggestive. It is significant that he focuses on rational forms of domination not by attacking science as such but by deconstructing those sciences in which human beings are both object and subject. These social, political, medical, and administrative sciences are deeply embedded in the power relations of modern societies. Undermining their claims to purity both provides a guiding thread for understanding modernity and theoretical support for new forms of political resistance on the terrain of knowledge. All this is certainly an advance over Marx. But although some of Marx and Marxists’ discussions of class and the state conform with Foucault’s notion of sovereignty, for Marx rational social arrangements are just as biased by the effects of power as they are for Foucault.

In support of this claim I must first explain what I mean by a “rational” social subsystem, and then revise the notion of bias accordingly. Let me begin then with the concept of “social rationality” which I have introduced to identify the peculiar character of many mostly modern institutions (Feenberg, 2008).

 It is obvious that no institution can be rational in exactly the same way as science and mathematics. Institutions are not held together by logical relations but by causal and symbolic forces that lack the rigor of experiment and equation. Nevertheless, procedures that bear a certain resemblance to those of science and mathematics operate in modern societies with tremendous effects on the whole social system. Organizations apply these procedures consciously and purposefully to gain control over social processes and the symbolic environment.

Social rationality in this sense depends on three such procedures. These are, exchange of equivalents, classification and application of rules, and optimization of effort and calculation of results. Each of these principles looks 'rational' as we ordinarily understand the term. The market, like calculation, is an exchange of equivalents. Bureaucracies resemble science in classifying objects and treating them uniformly under rules of some sort. And like science they measures their objects ever more carefully. Business, like technology, is based on optimizing strategies. Social life in our time thus appears to mirror scientific and technical procedures. This has consequences for the critique of social bias. That critique is based on another quite different kind of rationality, a dialectical rationality that enlarges contexts and connections in a process of interpretation. A discussion of that alternative form of rationality lies beyond the scope of this paper, however it will be constantly at work here.

We normally identify bias where prejudices, emotions and pseudo-facts of one sort or another influence judgments that ought to be based on objective standards. I call this “substantive bias” because it rests on a content of belief such as, for example, the idea that some races possess inferior intelligence. This concept of bias is an inheritance of the Enlightenment which aimed its critique at narrative legitimations of feudal and religious institutions. By contrast, the Enlightenment appealed to rational foundations, facts and theories unbiased by prejudice. There is no doubt that Enlightenment critique played and still plays an important role in emancipatory politics. However it has a significant limitation since it implies the neutrality and universality of systems such as technology, bureaucracy and markets that claim a rational foundation.

The critique of a rational system such as the market requires a different concept of bias. To explain that concept I need to distinguish it from romantic critique, which attributes substantive bias to rational systems and thereby denies the rationality of rationality as such. A similar critique is found in some postmodernist, feminist and STS theory. But whenever rationality is reduced to a non-rational origin such as Western or patriarchal ideology, or mere power relations, its special characteristics qua rational are overlooked.[1] This was not Marx's approach. He encountered it on the left of his day, for example in Proudhon, who titled his most famous book Property Is Theft. But if property really is theft, the coherence and survival of capitalism are incomprehensible. No social order can be based on simple plunder, certainly not one as complex and fragile as the capitalist system.

Marx's distinctive approach involved a very different style of critique based on the discovery of what I call “formal bias,” by which I mean a critique of the discriminatory effects of a rational order. I will argue that this is the methodological contribution of Marx that anticipates Foucault and contemporary STS.

Formal bias only becomes visible when rational systems are situated in their context. It is not a matter of prejudice and need not be defended with appeals to pseudo-facts or narrative myths because in this case the system itself objectifies the discriminatory principle. Criticism is silenced because the defender of the system can demonstrate its fairness on the terms of substantive critique. For example, a culturally biased test may discriminate effectively between populations, but those who design, administer and grade the test need not themselves be prejudiced for it to achieve a biased outcome.

Marx acknowledged the rational coherence of the market economy. But already in 1844 he cites “a contemporary economic fact. The worker becomes poorer the more wealth he produces" (Marx, 1963: 121). This fact suggests the hidden bias of the market which Marx sets out to explain as a consequence of the rational structure of capitalism. It is a difficult challenge. Habermas has succinctly summarized the problem Marx faced: "The institution of the market promises that exchange relations will be and are just owing to equivalence….The principle of reciprocity is now the organizing principle of the sphere of production and reproduction in itself" (Habermas, 1970: 97). In this passage Habermas explains the astonishing coincidence of mathematical equivalence and moral reciprocity in market relations. It is this equivalence that legitimates the market and makes it seem both natural and good.

Marx overcame this legitimating strategy in his theory of surplus value. I recall his argument here not to revive Marxist economics but as an example of a methodological innovation that has continuing validity and usefulness. In the ideal model of the capitalist economy that Marx derives from his bourgeois predecessors goods are paid for on the average at their value and their value consists in the labor required to produce them. Labor power itself is a good and the labor required to produce it is measured by the cost of food and other necessities. But because the capitalist owns the factory he has the power to set the length of the working day independent of the value of the labor performed within its compass. During the long working day workers produce goods worth more than the cost of their wages and so they enrich the capitalist to whom their product belongs. Meanwhile, the workers themselves remain at a mere subsistence level sufficient to reproduce their labor power.

Marx makes no reference to prejudices or pseudo-facts in this critique of capitalism. Surplus value is produced by the rational workings of the system itself. Property is not theft because labor is paid at its value. This is why Marx objected to early union demands for a fair wage. The problem is not with a specific rate of wages but with the structure of the market. However Marx's argument does effectively refute the normativity the market acquires when it is viewed as a pure exchange of equivalents, outside the context in which it actually functions as a mechanism of exploitation.

I want to turn now to the relation between Marx’s conception of formal bias and his critique of technology. Marx was not a rigorous technological determinist despite having written some famous passages in which he says that the “forces of production” determine the relations of production and all of social life. The bulk of his concrete discussions of technology concern the harm caused by industrial work. These passages seem to imply a critique of technology as such. But Marx rejected any such imputation and blamed the problems on the capitalist employment of machinery. But there are also a few passages in which technology is criticized for its specifically capitalist character. He writes for example that science "is the most powerful weapon for repressing strikes, those periodical revolts of the working class against the autocracy of capital" (Marx, 1906 reprint: 475). And further "it would be possible to write quite a history of inventions, made since 1830, for the sole purpose of supplying capital with weapons against the revolts of the working class" (Marx, 1906 reprint: 47-48). These passages suggest the contingency of science and technology on the interests of the capitalist class. Power/knowledge avant la lettre appears to be at work in Marx (Feenberg, 2002: chap 2).

The industrial revolution is the first time in history that basic economic production concerns members of the upper classes. Capitalists possess literate skills and access to scientific knowledge. At the same time they are in touch with the crafts in which lower-class people are engaged. Their familiarity with these two worlds of knowledge was one factor enabling them to restructure the labor process in order to eliminate costly craft labor and to introduce machines. This is the process of division of labor celebrated by Adam Smith in The Wealth of Nations. Our everyday sense of the word “technology” derives from this transition although of course earlier crafts also employed more or less sophisticated technical means.

Smith emphasizes the technical increase in efficiency, but there are other considerations of a political nature. As Andrew Ure wrote in 1835: “By the infirmity of human nature it happens, that the more skillful the workman, the more self-willed and intractable he is apt to become, and, of course, the less fit a component of a mechanical system, in which, by occasional irregularities, he may do great damage to the whole. The grand object therefore of the modern manufacturer is, through the union of capital and science, to reduce the task of his work-people to the exercise of vigilance and dexterity” (Ure, 1835: 18).

Competition drives the process of deskilling, but it would not be an effective economic strategy in the absence of the specific social relations of capitalism. Once labor becomes wage labor and its tasks are parceled out, production units no longer have a quasi natural character, rooted in community and family and supported by craft guilds and their traditions. The workers have no interest in production in both senses of the word, and even understanding the work plan becomes more and more difficult for those who implement it. The capitalist is thus technically necessary. Without his exercise of “power/knowledge” workers would slack off and coordination would break down. Foucault explains this aspect of factory work with explicit reference to Marx (Foucault, 1977: 175).

But Marx’s analysis goes further, explaining how, in Foucault’s words, power relations come to “function in a function.” Deskilling leads to mechanization. Once craft work is broken down into its simplest elements and each element assigned to a specific worker in a new division of labor, the potential role of machines in performing it becomes evident. Much of the capitalists’ role can be objectified in such machinery. Thus a social demand, in this case of capital, presides over the Industrial Revolution and orients much innovation throughout the 19th and 20th centuries.

Deskilling is such a general feature of invention over the last few centuries that it appears to be essential to progressive economic development. But in fact the history of capitalist enterprise shows just how contingent it is on peculiar social conditions and class conflicts, just as Marx supposed. What makes this difficult to perceive is the technically rational form taken on by these contingent developments. As a result old ideas about progress shared by Marxists and liberals alike prevailed until the 1970s when Harry Braverman’s pathbreaking book Labor in Monopoly Capitalism (1974) renewed Marx’s critique of deskilling.

David Noble followed with influential studies of the role of deskilling in American industrialization. Noble’s famous example of the automation of the machine tool industry has had a wide influence. Machine tools can be automated in two different ways. An early analog record/playback system was introduced by General Electric but it found no buyers because it still relied on craft workers to record a program. Management held out for digital systems that would translate directly from engineering drawings to machine movements, completely cutting craftsmen out of the loop. Noble’s argument exemplifies the workings of what constructivists call underdetermination. Management’s choice between these systems was ultimately decided by its ideological hostility to craft labor which was supported by a long tradition of management science, and not by neutral technical or economic reasons (Noble, 1984).

Noble does not claim that management got what it wanted nor that the sort of deskilling prophesized by Smith and Ure was actually achieved.[2] What is currently called “materiality” or the “agency” of things enters into his account. In fact, the digital systems could not be operated without help from skilled workers. Furthermore, the detailed working out of this example does not reinstate the humanistic notion of sovereignty criticized by Foucault. Like the Marxian sources on which it is based, Noble’s argument carefully situates human action in the context of social, technical and cognitive constraints and affordances.

The writings of Braverman, Noble and others in this tradition helped free Marxists from a naïve view of capitalist technological advance as a universal achievement. According to the new approach forces and relations of production co-construct each other. This term, “co-construction,” belongs to recent  constructivist technology studies which has rediscovered the Marxian idea of the interdependence of the social and the technical. In this field technology is conceived not as a pure product of inventive genius or as an application of science but as a “construction” of social actors. The technical underdetermination of artifacts leaves room for social choice between different designs that have overlapping functions but better serve one or another social interest. This means that context is not merely external to technology, but actually penetrates its rationality, carrying social requirements into the very workings of gears and levers, electric circuits and combustion chambers.

These concepts were introduced in an influential article by Trevor Pinch and Wiebe Bijker that made the connection between constructivism in science studies and a parallel approach to technology. They illustrated their argument with the history of the bicycle. As in Noble’s example of machine tools, so in this case two main designs were in competition. One of these designs looked a lot like bicycles today. It was relatively safe to ride but slow, useful only as a means of transportation. The other had a high front wheel and was faster but less stable. It appealed to young sportsmen who liked to race. The different designs thus corresponded to the requirements of the different social actors. The triumph of the low wheelers resulted from the introduction of air filled tires to reduce vibration. When these were tried out in bicycle races, the low wheelers proved fast as well as stable and soon became the preferred design. “Closure” was achieved, but the outcome was contingent and not the result of “progress” in the sense in which that term is understood in a deterministic context. Pinch and Bijker comment: “[T]he different interpretations by social groups of the content of artifacts lead by means of different chains of problems and solutions to different further developments…” (Pinch and Bijker, 1987: 42).

Pinch and Bijker’s key point is the influence of the social on “the content of the artifact itself” and not merely on such external factors as the pace of its development or its use (Pinch and Bijker, 1987: 42). This appears to parallel a similar constructivist relativism in science studies but the parallel is misleading. Although similar in many respects, as constructivists argue, facts and artifacts are quite different in the most important respect. The “hardening” of natural scientific facts by experiment and replication yields a “content” that is far more compelling than a successful technological design. Of course the elements of engineering are often firmly established by scientific methods and long empirical experience, but they are a far cry from a finished artifact. They must be combined in contingent combinations and it is in the course of this process that the “interpretative flexibility” of technology becomes evident. The underdetermination of the outcome is radical and obvious to technologists. Its demonstration requires no subtle epistemological arguments. Foucault’s caution in restricting his study to sciences such as psychiatry and criminology reflected an appreciation of their similarly “soft” status.

As the deskilling contest clearly shows, the social construction of technical artifacts is an intervention into the lives of their users. The “scripts” inscribed in the artifacts govern their usage and hence a large part of social behavior, in some cases even users’ way of life. Users are configured by artifacts but in turn influence their design. Interactions played out in the course of the diffusion of technologies co-construct both human beings and the technologies themselves (Woolgar, 1991; Akrich, 1992; Oudshoorn and Pinch, 2003). I have studied several cases in which these complex relations between users and designers played a key role. In my study of the French Minitel I showed how the very meaning of the system was transformed by hackers who turned an information network designed to rationalize French society into a communication network for the exchange of instant messages between users seeking dates (Feenberg, 1995: chap. 7). The original program which was designed to “configure” the users as more efficient members of an “information society” was subverted as the users resignified the system as a place of human encounter.

This very brief sketch of constructivist technology studies does not do justice to the rich body of research in the field but it gives a hint of its relevance to issues of great social and political consequence. The central achievement of these theories is the liberation of technology studies from deterministic models and the introduction of a hermeneutic perspective. The new question of technology is not about efficiency but about meaning.

From this standpoint, the history of deskilling as Marxists analyze it is a hermeneutic contest between two different visions of work and the worker, one promoted by capital and the other by workers. Deskilling involves the realization of the capitalist interpretation of work in the technical specifications of factory equipment. A new type of society is built around this transformation.

I have introduced the term “technical code” to refer to the general rules for translating such meanings between the language of social actors and the technical languages of the day. In earlier work I explored this concept in relation to child labor, a consequence of deskilling which led to the construction of machines sized for operation and maintenance by children. When factory owners were confronted with child labor legislation, they protested that the “imperatives of technology” required the small workers. The meaning of labor and hence of technology was at stake in the expulsion of children from the factory (Feenberg, 1991: 127-131).

Although such arrangements tend to be exported to China these days, we are surrounded by other examples of technical codes. Environmentalist concern with climate change is translated into the specifications of engines and building codes. To automotive engineers, safety literally means seatbelts, airbags, electronic skid control, and so on. In Bruno Latour’s vocabulary, each language is employed to mobilize different networks; ordinary language mobilizes human actors, including the general population and its political leaders. Among those so mobilized are technical experts whose technical discourse mobilizes the non-human actors required to carry out programs such as fighting climate change or building safer cars. Deskilling is a technical code expressed on the one hand in discursive form as an ideological preference for the replacement of skilled labor by machinery, and on the other hand, expressed in technical specifications that determine machines with precisely this function. The technical code of capitalism translates the requirements of control from above into technology (Feenberg, 1999: 97-89).

Marx hints at the possibility of a socialist technical code that would liberate intelligence and skill. In the Grundrisse, for example, he writes that labor under socialism must be "of a scientific and at the same time general character, not merely human exertion as a specifically harnessed natural force, but exertion as subject, which appears in the production process not in a merely natural, spontaneous form, but as an activity regulating all the forces of nature" (Marx, 1973: 612). This conception of socialist labor explains why, in “The Critique of the Gotha program,” Marx anticipates a day when work will be “life’s prime want” (Marx and Engels, 1972: 388; Feenberg, 2002: chap. 6).

This implies a very different evolution of technology under socialism. Presumably the “assembled producers,” as Marx called the empowered workers, would make different technical choices from the capitalists. A different technical code would be imposed by workers, who would appear as actors in the technical domain. Although Marx left all this rather vague, later Marxists have developed the argument for a democratic technical system while also addressing the limitation of Marx’s formulation to a proletariat that no longer exists in the form in which he knew it.

I want now to address the development of the politics of technology critique in more recent times. I have tried to show that Marx anticipated Foucault’s essential point, namely, that a critique of modern civilization must include rationality in its purview. The structure of wage labor and technology requires just such a critique. But these systems are no longer confined to capitalist enterprise as they were in Marx’s day. In the 20th century they spread to state institutions and eventually to communist societies. This shows that the factory was simply the initial site of a more general technical politics. The first place in which technology was widely deployed was the factory and so it was there that technical resistances first manifested themselves. But today technology is involved in every aspect of life, including domains unimagined by Marx such as leisure, education and even dating as we saw with the Minitel. Once technology spreads over the whole surface of society, a much wider range of technical struggles emerge, as is clear from the contemporary politics of the environment, medicine, and computerization.[3] Foucault’s critique of social knowledge was one of several theoretical initiatives inspired by this new situation.

The multiplication of scenes of struggle shows that the bias of technological rationality is due not to ownership but to what I call the “operational autonomy” of the capitalist and his administrative successors. By operational autonomy I mean the freedom of the owner or manager to make independent decisions regardless of the views or interests of subordinate actors and the surrounding community. Operational autonomy positions management in a technical relation to the world, safe from the consequences of its own actions. There it is able to reproduce the conditions of its own supremacy at each iteration of the technologies it chooses and commands. Technocracy is an extension of such a system to society as a whole in response to the spread of technology and management to every sector of social life. A self-perpetuating dynamic emerges. We could employ Thomas Hughes’ term and call this a kind of “momentum” acquired by modern organizations as they expand and develop.

Unfortunately, the socialist movement remained fixated on ownership and expected miracles from nationalizations without democratization of technical relations. The neutrality of technology became doctrine among Marxists as it was among capitalists and direct transfers of the most oppressive Western technological designs formed the underpinnings of industrialization in the Soviet Union. The failure to place the problem of operational autonomy at the center of the theory of transition led to a catastrophic centralization of power in Russia, China and other communist countries. By the 1970s this approach was discredited among most Marxist theorists in the West. Labor process theory and various forms of neo-Marxism restored a more sophisticated understanding of Marxian social theory that converged in certain respects with Foucault’s non-Marxist critique of rationality.

Revisionist Marxism in the 1960s and 70s developed a radical critique of modern civilization as a whole. The most famous representative of this trend was Herbert Marcuse. He worked out the complex of critical ideas that have by now become clichés: the stabilization of the “one-dimensional” system through media propaganda, technocratic ideology, privatization, consumerism, and the displacement of surplus aggression onto racial or foreign scapegoats.

In addition, Marcuse challenged blind faith in “progress” with a radical critique of technological rationality. In the dominant ideology technology was represented as a pure application of knowledge of nature, above political and social differences. The rational character of technology served as its alibi, freeing it from responsibility and placing it beyond controversy. Marcuse contested this image of neutral, value free technology. Modern technology does differ from traditional crafts, bound to the service of specific culturally secured values. But its liberation from such values simply places it in the service of the dominant actors. This aspect of Marcuse’s argument leads in my own work to the concept of formal bias.

Marcuse was less concerned with the problem of authority that preoccupied Foucault than with the intrinsic limitations of modern scientific-technical rationality. His argument reprised certain themes of the romantic critique of reason, such as its suspicion of quantification, albeit in a political form. Marcuse contrasted a science and technology of domination with an alternative realization of rationality respectful of the developmental potential of nature and human beings.  Such an alternative, Marcuse argued, would be rooted in the imaginative understanding of the world and the sense of beauty, themes that resonated with the counter-culture of the day (Feenberg, 2005: chap. 5).

Marcuse’s argument thus culminates in a positive alternative. He argued for "rupture with the continuum of domination, the qualitative difference of socialism as a new form of way of life, not only rational development of the productive forces, but also the redirection of progress toward ending of the competitive struggle for existence, not only abolition of poverty and toil, but also reconstruction of the social and natural environment as a peaceful, beautiful universe: total transvaluation of values, transformation of needs and goals. This would mean not to regress in the technical progress, but to reconstruct the technical apparatus in accordance with the needs of free men, guided by their own consciousness and sensibility, by their autonomy" (Marcuse, 1970: 280).

Marcuse's ecstatic projection sounds shockingly utopian today, but it provides a vision of a very different technological future, and some sort of vision is needed if we are ever to see the fundamental changes required by an adequate response to the environmental crisis. Marcuse’s influence reached a highpoint in the 1970s when the counter-culture promised a radical break with the American Way of Life. The activists of this time called for the pursuit of happiness in other than commodity forms. Marcuse’s formulation of the critique of consumerism is significant for rejecting appeals to individual self-denial and instead emphasizing the necessity of restructuring of the technological base of the system. He argued that the civilizational project that shapes modern technology can be replaced by another way of life based on a technology that respects nature. The disappearance of the counter-culture has not refuted this message which is reiterated today more moderately by the environmental movement. “Demythologized” for contemporary purposes, Marcuse’s argument can be rephrased in terms of the underdetermination of technology, which opens the possibility of alternative modernities.

This historical background explains the decline of the positivist rationalism and technocratic determinism that achieved intellectual hegemony in the English speaking world after the Second World War. The dramatic events of the 1960s, followed by the long slow rise of environmentalism, eroded the hegemonic certainties of the post-war period. The possibility of alternatives, repeatedly illustrated by successful environmental regulation, sapped the authority of technocratic experts who claimed to know the one best way to do things. Rationality was no longer one but multiple, at least in its socially concrete realizations, and so subject to politics rather than over-ruling it in the name of a superior truth.

The changed situation of rationality explains the credibility of the new social studies of technology that emerged to prominence in the 1980s. By the end of the century academic study of technology had abandoned determinism and instrumentalism in favor of various constructivist alternatives. This depoliticized version of a critique of technology introduced important methodological innovations as we have seen. But constructivist social science was remarkably inhibited when it came to studying itself. If it had violated this taboo it would have discovered the background discussed here earlier and especially the importance of social struggles over technology in giving its theoretical innovations significance beyond a specialized audience.

This inhibition no doubt explains the shock of the so-called “science wars,” in which a rearguard defense of typical positivistic and technocratic theses confronted constructivists many of whom were convinced, like the scientists they studied, that they were not responsible for the politics of their research (Gross and Levitt, 1994). The stakes were raised when attacks came from the left, the natural home of most constructivist scholars. For example, Meera Nanda argues that post-modernism and constructivism are operationalized by Hindu nationalists. She writes, “Neo-Hinduism and Hindutva are reactionary modernist movements, intent on harnessing a mindless and even dangerous technological modernisation for the advancement of a traditionalist, deeply anti-secular and illiberal social agenda. Nevertheless, they share a postmodernist philosophy of science that celebrates the kind of contradictory mish-mash of science, spirituality, mysticism and pure superstition that passes as ‘Vedic science’” (Nanda, 2004; see also Nanda, 1996). Whether these attacks on constructivism were entirely justified is a question I will not address, but the weak response in the STS community revealed real problems.

It had evidently not occurred to constructivists that the relativism that made their theory effective against technocracy made it useless against fundamentalism and traditional forms of racism and sexism. The problem was made insoluble by the identification of science and technology in a supposedly unified “technoscience.” Unless science and technology are distinguished there is no way to avoid the contradiction between a progressive, democratic argument against the subjective bias of fundamentalism and the parallel argument against the formal bias of technocracy.[4] To make matters worse, the STS community was at first reluctant to make its anti-technocratic argument explicit. What Wiebe Bijker once called the “academic detour” had led far from the main political road as feminist critics noted (Wajman, 2004: 126). It took a while to find the way back but eventually students of technology such as Bruno Latour and Bijker himself engaged more openly with the core political issue, the relation of technology to democracy (Latour, 1999; Bijker, 1996).[5] The theme of democratization of technology, an obvious political implication of the thesis of underdetermination, thus began to receive the attention it deserves. If this theme is pursued, the encounter of technology studies and political theory may bring about a major reconfiguration of the social sciences.

As Latour has argued, the exclusion of technology from the social scientific concept of society is untenable. But once technology enters the picture, the issue of rationality appears in a new guise. It was Weber who introduced the concept of rationalization to explain many of the processes Marx had earlier identified as central to modernity. But whereas in Marx such processes were conceived as potentially multiple—capitalist or socialist—Weber argued that they were the same for all modern societies. Now, with the new focus on technology, we have a basis for questioning Weber’s simplification. Because of the underdetermination of technology, rationalization must reproduce the multiplicity of the social. There is no single homogeneous outcome to technological development and so none to social development either. The future, that seemed closed by the certainties of social science, opens up anew. The question of democratization concerns the form in which social influences on development will be exercised and institutionalized in the rationalized domains.

This paper has concerned critical strategies for addressing the resistance of rationality to rational critique especially with respect to technology. When an action is rationally justified, how can reason deny its legitimacy? Foucault addressed this problem in his theory of power/knowledge. Marx anticipated that approach in his critique of the “social rationality” of the market and technology. Marx’s discussion of the length of the working day got around the silencing effect of social rationality with something very much like the concept of underdetermination. There are hints of a critique of capitalist technology in his writings as well. I have introduced the distinction between substantive and formal bias to explain the difference between critiques of prejudice and Foucault and Marx’s style of critique. Rather than identifying factual errors, they enlarge the context of explanation of the workings of underdetermined rational systems in such a way as to reveal hidden biases.

In the 1970s, Marxists rediscovered Marx’s critique against a background of rising political and cultural struggles. They argued that the design of technologies such as the assembly line is shaped by capitalism. At about the same time neo-Marxists and post-structuralists demanded radical changes in the technological rationality of advanced societies. The spread of technology into every nook and cranny of social life provided a context in which the thought of Foucault and Marcuse achieved widespread popularity. The subsequent political reaction reduced the influence of their radical theories but also saw the emergence of an environmental movement that focused attention on technology in new ways. Other social movements appeared in medicine and around computerization that sharpened the technological focus. In this changed context academic study of technology took new paths. The concept of underdetermination was finally formulated clearly in contemporary science and technology studies, but without explicit political purpose. Nevertheless, this revision of the academic understanding of technology contributes to weakening technocratic rationales for public policy. A new era of technical politics has begun.






Akrich, Madeleine, ‘The de-scription of technical objects’, in Wiebe E. Bijker and John Law, eds, Shaping technology/Building society: Studies in sociotechnical change. Cambridge MA: MIT Press, 1992. pp. 205-224.

Bijker, Wiebe E. (1996). “Democratization of Technology, Who are the Experts?” http://www.angelfire.com/la/esst/bijker.html (consulted April 5, 2008).

Braverman, Harry (1974). Labor in Monopoly Capitalism: The Degradation of Work in the Twentieth Century. New York: Monthly Review Press.

Feenberg, Andrew (1991). Critical Theory of Technology. New York: Oxford.

Feenberg, Andrew, 1995.  Alternative Modernity: The Technical Turn in Philosophy and Social Theory. Los Angeles: University of California Press.

Feenberg, Andrew (1999). Questioning Technology. New York: Routledge.

Feenberg, Andrew (2002). Transforming Technology. New York: Oxford University Press.

Feenberg, Andrew (2005). Heidegger and Marcuse: The Catastrophe and Redemption of History. New York: Routledge.

Feenberg, Andrew (2008). “From the Critical Theory of Technology to the Rational Critique of Rationality,” Social Epistemology, Vol. 22, no. 1, Jan.-March 2008, pp. 5-28.

Foucault, Michel (1977). Discipline and Punish, A. Sheridan, trans. New York: Pantheon.

Foucault, Michel (1980). Power/Knowledge, C. Gordon, trans. New York: Pantheon.

Foucault, Michel (1988). The Final Foucault, J. Bernauer and D. Rasmussen, eds. Cambridge, Mass.: MIT Press.

Gross, Paul, and Levitt, Norman (1994). Higher Superstition: The Academic Left and Its Quarrels with Science. Baltimore: Johns Hopkins University Press.

Habermas, Jurgen (1970). “Technology and Science as Ideology,” in Toward a Rational Society, J. Shapiro, trans. Boston: Beacon Press.

Latour, Bruno (1999). Politiques de la nature: Comment faire entrer la science en démocratie. Paris: La Découverte.

MacKenzie, Donald (1996). Knowing Machines. Cambridge, Mass.: MIT Press.

Marcuse, Herbert (1970). “Re-Examination of the Concept of Revolution,” in All We Are Saying, A. Lothstein, ed. New York: Capricorn Books.

Marx, Karl (1963). Karl Marx: Early Writings, T. Bottomore, ed. London: C.A. Watts.

Marx, Karl and Engels, Friedrich (1972). The Marx-Engels Reader, R. Tucker, ed. (New York: Norton.

Marx, Karl (1973). Grundrisse. Baltimore: Penguin.                

Nanda, Meera (1996) “The Science Question in Postcolonial Feminism,” in Annals of the New York Academy of Sciences 775 (1) , 420–436.

Nanda, Meera (2004). "Postmodernism, Hindu nationalism and `Vedic science.'" Frontline, Volume 20 - Issue 26, December 20, 2003 - January 02. Consulted April 5, 2008 at http://www.frontlineonnet.com/fl2026/stories/20040102000607800.htm

Noble, David (1984). Forces of Production. New York: Oxford University Press.

Oudshoorn, Nelly and Trevor Pinch, eds., How users matter: The co-construction of users and technology. Cambridge MA: MIT Press, 2003.

Pinch, Trevor, Bijker, Wiebe (1987). "The Social Construction of Facts and Artefacts," in Bijker, Wiebe, Hughes, Thomas, Pinch, Trevor, eds., The Social Construction of Technological Systems. Cambridge, Mass.: MIT Press.

Ure, Andrew (1835). The Philosophy of Manufactures. London: Charles Knight.

Wajcman, Judy (2004). Technofeminism. Cambridge, UK: Polity Press.

Woolgar, Steven, ‘Configuring the user: The case of usability trials’, in John Law, ed, A sociology of monsters. London: Routledge, 1991.

[1] The risk of regression to which such approaches are exposed is very real: Christian fundamentalists have been able to appeal to science studies in defense of their right to teach “intelligent design” in the schools.

[2] It is true that new skills often arise as old ones are destroyed, but from the standpoint of contemporary actors, this is irrelevant. Managers seeking to cut labor costs and workers defending their jobs and way of life confront each other in the present regardless of what may happen in the future.

[3] I have offered examples of such struggles in several of my books: medicine (Feenberg, 1995: chap. 5), computerization (Feenberg, 1995: chap. 7), educational technology (Feenberg, 2002: chap. 5), environmentalism (Feenberg, 1999: chap. 3).

[4] Of course the distinction is not always sharp  but like many other fuzzy distinctions it is nonetheless real and consequential. See (Feenberg, 2002: 171-174).

[5] I have attempted to develop a quasi-constructivist political theory of technology in (Feenberg, 1999: Part II)