Loading field map
Is a nuclear power plant a political actor in its own right, or is it merely a tool that democracies can choose to regulate? Does the internet reshape human cognition, or do human intentions steer its development? These questions have animated the philosophy of technology for over a century, and the answers have shifted dramatically. At the heart of the subfield lies a persistent tension: is technology an autonomous force that remakes society, or is it a socially shaped product that reflects human choices? Different frameworks have taken up this tension, each responding to the limitations of its predecessors, and the result is a rich, evolving conversation about how to understand the relationship between technical systems and social life.
The earliest systematic frameworks in the philosophy of technology tended to see technology as a driving force that shapes society from the outside. Marxist Theory of Technology, rooted in the work of Karl Marx in the mid-19th century, argued that the means of production—the tools, machines, and infrastructure of an era—determine the relations of production, including class structure and political power. For Marx, the steam mill gave rise to industrial capitalism, and the hand mill gave rise to feudal society. Technology was not neutral; it was the engine of historical change. Yet Marx also saw technology as a site of struggle: capitalists and workers could fight over its control and design, a point that later frameworks would develop further.
In the mid-20th century, Technological Determinism emerged as a more explicit and general claim: technology develops according to its own internal logic, and society must adapt to its imperatives. Thinkers such as Jacques Ellul and Langdon Winner (in his early work) argued that technical efficiency becomes the dominant value in modern societies, overriding political, ethical, or cultural considerations. Ellulian Technique, introduced in Ellul's 1954 book The Technological Society, went further: technique is not just machines but a whole mode of rational, efficient thinking that colonizes every domain of life, from economics to education to warfare. For Ellul, technique is an autonomous, self-augmenting force that humans cannot control.
Around the same time, Martin Heidegger offered a different but equally powerful determinist account. Heideggerian Enframing, presented in his 1954 essay "The Question Concerning Technology," argued that modern technology is not a neutral instrument but a way of "revealing" the world as a standing reserve—a set of resources to be ordered, optimized, and exploited. Enframing is an ontological condition: it shapes how humans perceive reality itself, reducing nature and even human beings to calculable inputs. Unlike Ellul, Heidegger did not see technology as a historical force that could be resisted through political action; instead, he called for a meditative thinking that could open up a different relationship to Being. These four determinist frameworks—Marxist, technological, Ellulian, and Heideggerian—shared a conviction that technology is not a passive tool, but they disagreed on its ultimate source (economic relations, technical logic, or ontological disclosure) and on whether resistance was possible.
By the 1980s, a growing number of scholars found determinism too sweeping and too pessimistic. They wanted to open the "black box" of technology and examine how specific artifacts and systems are actually designed, adopted, and shaped by social groups. Technological Politics, articulated most influentially by Langdon Winner in his 1980 article "Do Artifacts Have Politics?," argued that technical objects can embody political arrangements. Winner's famous example was the low-hanging overpasses on Long Island, which he claimed were deliberately designed by Robert Moses to prevent buses (and thus poor and Black residents) from reaching public beaches. Technology, in this view, is not autonomous but is infused with political intentions from the start.
Large Technical Systems (LTS), developed by Thomas Hughes and others from 1983 onward, studied the historical development of massive infrastructures like electrical grids, telephone networks, and railroads. Hughes showed that these systems are not simply imposed by technology; they emerge through a complex interplay of inventors, engineers, managers, financiers, and regulators. LTS scholars emphasized the "momentum" of large systems—once established, they become difficult to redirect—but they also showed that systems are built through human choices at every stage. This was a more nuanced determinism: systems have inertia, but they are not autonomous.
The most explicit rejection of determinism came with the Social Construction of Technology (SCOT), introduced by Trevor Pinch and Wiebe Bijker in 1984. SCOT drew on the sociology of scientific knowledge to argue that technological artifacts are "socially constructed": different social groups interpret the same artifact differently, and the final design is the outcome of negotiation, conflict, and closure. For example, the early bicycle went through many designs (high-wheelers, safety bicycles, tricycles) before settling into its modern form, and that settlement reflected the interests of different users—sporting young men, women, elderly riders—not just technical efficiency. SCOT focused on the micro-level of specific artifacts and the interpretive flexibility of their meanings.
Social Shaping of Technology (SST), developed by Donald MacKenzie, Judy Wajcman, and others from 1985 onward, shared SCOT's anti-determinism but operated at a broader level. SST examined how economic, political, and cultural forces shape entire technological trajectories, not just individual artifacts. Where SCOT asked how a specific design was negotiated, SST asked why certain technologies (like nuclear power or the internal combustion engine) were pursued while others were abandoned. SST also paid more attention to the role of institutions, markets, and state power. The two frameworks coexisted and complemented each other: SCOT provided fine-grained case studies of artifact design, while SST offered a macro-level account of technological change.
Even as constructivism opened the black box, some scholars felt it remained too human-centered. Actor-Network Theory (ANT), developed by Bruno Latour, Michel Callon, and John Law from 1986 onward, insisted that non-human entities—machines, animals, texts, natural forces—are also actors that shape outcomes. For ANT, a scientist's laboratory, a speed bump, or a scallop-fishing net is not a passive object but a participant in networks that produce effects. ANT rejected the distinction between "social" and "technical" factors; instead, it analyzed heterogeneous networks of human and non-human actors. This was a radical departure from SCOT and SST, which still treated society as the primary shaping force. ANT argued that agency is distributed and symmetrical: we should explain the social and the technical in the same terms.
Postphenomenology, developed by Don Ihde from 1990 onward, offered a different kind of relational approach, rooted in phenomenology rather than sociology. Ihde drew on Heidegger but rejected his monolithic concept of Enframing. Instead, Postphenomenology examined how specific technologies mediate human perception and experience. A microscope, a hearing aid, or a smartphone does not simply impose a worldview; it transforms the user's relationship to the world in particular ways—amplifying some aspects, reducing others. Postphenomenology focused on the embodied, situated experience of using technology, and it emphasized the variability of mediation: the same technology can be used differently in different contexts. This was a more fine-grained alternative to Heideggerian Enframing, which had treated all modern technology as a single, oppressive mode of revealing.
By the 1990s, many scholars wanted to move beyond describing how technology is shaped and toward actively shaping it for ethical and political ends. Critical Theory of Technology, developed by Andrew Feenberg from 1991 onward, synthesized insights from Marx and Heidegger while rejecting their pessimism. Feenberg argued that technology is not autonomous but is a "field of struggle" where different social groups can contest its design and purpose. He drew on the Frankfurt School's critical theory to argue that modern technology embodies a narrow, instrumental rationality, but that democratic interventions—such as worker participation in design or user-led innovation—can transform it. Critical Theory of Technology thus preserved Marx's emphasis on social conflict and Heidegger's critique of instrumental reason, but it added a hopeful, democratic politics.
Feminist Technology Studies, emerging around the same time, brought gender analysis to the study of technology. Scholars such as Judy Wajcman, Cynthia Cockburn, and Ruth Schwartz Cowan showed that technologies are not gender-neutral: they are designed, marketed, and used in ways that reflect and reinforce gender hierarchies. For example, the microwave oven was marketed to women as a tool for domestic efficiency, while the personal computer was initially coded as a male hobby. Feminist Technology Studies did not simply add gender as a variable; it argued that gender relations are constitutive of technological development. This framework challenged both determinism (which ignored gender) and earlier constructivism (which had not systematically examined gender).
Constructive Technology Assessment (CTA), developed in the Netherlands from 1992 onward, was a more practical, policy-oriented framework. CTA aimed to bring social and ethical considerations into the design process early, before technologies become locked in. It involved stakeholders—users, citizens, experts—in dialogue about emerging technologies, with the goal of shaping development in socially desirable directions. CTA was explicitly interventionist: it did not just analyze technology-society relations but sought to improve them through democratic deliberation.
In the 2000s and 2010s, new frameworks emerged that addressed the scale and complexity of contemporary technological systems. Sociotechnical Transitions, developed by Frank Geels, Jan Rotmans, and others from 2002 onward, studied how entire socio-technical systems—such as energy, transportation, or water supply—undergo fundamental change. This framework drew on LTS and evolutionary economics to analyze the interplay between niche innovations (like solar panels), existing regimes (like fossil-fuel-based electricity), and broader landscape pressures (like climate change). Sociotechnical Transitions offered a multi-level perspective that explained both stability and radical change, and it became influential in sustainability transitions research.
Value Sensitive Design (VSD), articulated by Batya Friedman and others from 2003 onward, provided a methodology for incorporating ethical values—such as privacy, autonomy, and justice—into the design of information systems and other technologies. VSD was a direct response to the recognition that technologies embody values, but it went further by offering concrete design methods: conceptual investigations (identifying relevant values), empirical investigations (studying how users experience those values), and technical investigations (designing features that support them). VSD coexisted with CTA and Responsible Research and Innovation, but it focused more on the design phase and less on broader governance.
Sociotechnical Imaginaries, introduced by Sheila Jasanoff and Sang-Hyun Kim in 2009, examined how collective visions of desirable futures shape technological development. Imaginaries are not just individual hopes but shared, institutionally stabilized visions—such as the American dream of a nuclear-powered future or the European vision of a sustainable energy transition. This framework connected the study of technology to political culture, showing that what societies imagine as possible and good influences which technologies are pursued. Sociotechnical Imaginaries complemented SCOT and SST by adding a cultural and political dimension: it was not just about interests and negotiations but about shared meanings and aspirations.
Responsible Research and Innovation (RRI), developed from 2012 onward, emerged as a framework for governing emerging technologies like nanotechnology, synthetic biology, and artificial intelligence. RRI built on CTA and VSD but added a stronger emphasis on anticipation, reflexivity, inclusion, and responsiveness. It called for researchers and innovators to consider the potential impacts of their work early, to engage diverse stakeholders, and to adapt their practices in light of public values. RRI was a normative framework that aimed to make innovation processes more democratic and ethically accountable.
Today, no single framework dominates the study of technology and society. The field is characterized by productive pluralism. Most contemporary frameworks agree on a few core points: technology is not a neutral instrument; it is shaped by social, political, and cultural forces; and it in turn shapes those forces in a co-constitutive relationship. They also agree that determinism, in its simple form, is inadequate. However, they disagree on several key issues. One major disagreement concerns agency: Actor-Network Theory insists that non-humans are actors, while most other frameworks reserve agency for humans and social structures. Another disagreement is about scale: SCOT and Postphenomenology focus on micro-level interactions with specific artifacts, while Sociotechnical Transitions and Sociotechnical Imaginaries analyze macro-level systems and cultural visions. A third disagreement is about normativity: frameworks like VSD and RRI are explicitly prescriptive, aiming to improve design and governance, while frameworks like SCOT and ANT are primarily analytical, seeking to understand how technology-society relations work. These disagreements are not signs of weakness; they reflect the richness of the subject and the different questions scholars ask. The field continues to evolve, with new frameworks emerging as technologies themselves change, but the central tension between technology as autonomous force and technology as social product remains the enduring thread that ties the conversation together.