Actor-Network Theory Analysis
- Introduced by
- Bruno Latour, Michel Callon, John Law
- Year
- 1984
- Type
- Material-semiotic theory and analytic method
- Family
- Sociology of translation
What does it do?
Actor-Network Theory analysis treats society and technology as a single woven fabric, mapping how heterogeneous human and non-human actors—engineers, scallops, documents, machines, regulators—are linked into networks through a process of translation.
Rather than explaining technical outcomes by appeal to pre-given social categories, ANT follows the actors themselves and describes how durable arrangements are assembled, stabilised, and sometimes undone.
The intuition behind the method
Imagine trying to explain why a new transport card succeeded. ANT refuses to say 'because society wanted it.' Instead it follows every participant—the card chip, the turnstile, the commuters, the transit authority, the standards document—and asks how each was persuaded or forced into line. The card only 'works' because a whole network of allies holds together; pull out the chip standard or the commuters' habits and the network frays. Power, in this view, is not possessed but is the effect of how many allies you have enrolled and how firmly they are tied in.
History
Actor-Network Theory emerged at the Centre de Sociologie de l'Innovation in Paris in the early 1980s through Michel Callon, Bruno Latour, and the British sociologist John Law. Callon's 1984 study of scallop domestication in St Brieuc Bay codified the sociology of translation, while Latour's laboratory studies and later Reassembling the Social (2005) elaborated the material-semiotic programme. ANT both extended and broke from the social constructivism of the period, and despite Latour's own ambivalence about the label it became one of the most influential frameworks in science and technology studies.
When to use it
Use ANT analysis when you want to explain a sociotechnical outcome—the success or failure of an innovation, the stabilisation of a standard, the workings of a scientific fact—without reducing it to either purely social or purely technical causes. It suits studies where non-human elements visibly shape action and where you can gather rich traces of how associations were built. ANT assumes generalised symmetry between humans and non-humans and a willingness to describe rather than to explain by hidden structures. It is less appropriate when you need causal generalisation across many cases, when power asymmetries should be theorised in advance, or when ethnographic access to the network's making is unavailable.
Limitations and considerations
- Its descriptive, anti-explanatory stance frustrates readers who want causal generalisation or prediction.
- Symmetry between humans and non-humans is philosophically contested and can flatten morally important asymmetries of power and agency.
- Findings depend heavily on which actors the analyst chooses to follow, raising questions of completeness and selection.
- It offers limited purchase on enduring structural inequalities that pre-exist any particular network.
Source conditions do not replace independent scientific review or validation by an analysis engine.
Strengths
- Dissolves the social/technical and macro/micro dichotomies, letting analysts describe innovation as a single heterogeneous process.
- Grants analytic voice to non-human actors—devices, standards, organisms—that conventional social science treats as inert background.
- Provides a precise vocabulary (translation, enrolment, obligatory passage point, black box) for narrating how networks are built and stabilised.
- Travels across science, technology, organisations, markets, and infrastructure, making it a flexible comparative lens.
Applications
- Explaining the success or failure of innovations by tracing how allies were or were not enrolled.
- Analysing how scientific facts are stabilised through networks of instruments, papers, and colleagues.
- Studying large infrastructures and standards as heterogeneous networks held together by translation work.
- Investigating market devices and economic arrangements as assemblages of calculative tools and agents.
- Mapping environmental and health controversies where organisms, pollutants, and instruments act alongside people.
How it works
How it works
Begin by following the actors—human and non-human—without deciding in advance which are 'social' and which 'technical'. The analyst suspends a priori distinctions and lets the field define who and what matters, granting symmetry to people, devices, texts, and natural entities as potential actants.
Document the concrete associations linking actors: who speaks for whom, which device delegates which action, how a document carries an intention across distance. The network is described as the sum of these traceable connections rather than as a hidden structure behind them.
Analyse translation through Callon's four moments—problematisation (defining the problem and actors), interessement (locking actors into roles), enrolment (negotiating those roles), and mobilisation (rendering enrolled actors into spokespersons). Translation is how one actor makes itself indispensable to others.
Identify the obligatory passage point—the node every actor must pass through to achieve its goals—which a focal actor constructs to channel and discipline the network around its own programme of action.
Assess the durability and reversibility of the assembled network: how black-boxed and irreversible it has become, where counter-enrolments or dissidence threaten it, and whether the arrangement holds or dissolves once scrutiny stops.
Common pitfalls
- Smuggling pre-given social categories (class, gender, the state) back in as explanations after claiming to follow actors symmetrically.
- Treating 'follow the actors' as licence for unbounded description with no analytic stopping rule.
- Attributing intentional agency to objects rather than the more careful claim that objects make a traceable difference to action.
- Ignoring the actors who were silenced or never enrolled, mistaking a stabilised network for the whole story.
Frequently asked questions
Is ANT a theory or a method?
Both, uneasily. ANT supplies an ontology (a material-semiotic, relational view of the world) and an analytic sensibility ('follow the actors', grant symmetry, describe translation) rather than a fixed procedure. Latour himself called it more a method or even an 'infra-language' than a predictive theory. In practice researchers operationalise it through ethnography, document tracing, and controversy mapping.
What does 'symmetry' mean in ANT?
Generalised symmetry is the methodological rule of using the same vocabulary for humans and non-humans and not deciding in advance that social factors explain technical ones (or vice versa). It does not claim objects are people; it claims that the analyst should not pre-sort actors into active humans and passive things, because devices, texts, and organisms demonstrably make a difference to how networks act.
How is ANT different from SCOT?
The Social Construction of Technology (SCOT) explains artefacts via the interpretations of pre-defined relevant social groups and tends to keep humans central. ANT extends agency to non-humans, dissolves fixed social groups in favour of traceable associations, and focuses on translation. SCOT asks which social groups shaped the artefact; ANT asks how the whole heterogeneous network—including the artefact—was assembled.
Related methods
Sources
Latour, B. (2005). Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford University Press.
ISBN: 9780199256051Copied.Callon, M. (1984). Some elements of a sociology of translation: domestication of the scallops and the fishermen of St Brieuc Bay. The Sociological Review, 32(1_suppl), 196-233.
Open DOI referenceCopied.