Actor–Network Theory

What has to come together before a message can act?

Suppose a local newsroom publishes a flood alert. A reporter gathers residents’ accounts. A water-level gauge supplies a reading. A spreadsheet converts measurements into categories. An editor approves a headline, a publishing system creates the page, and a mobile notification directs readers towards it.

The alert’s journey continues. A weak signal delays delivery. A cropped screenshot removes the time stamp. A community volunteer translates the instructions into Bengali. Residents compare the message with the water outside their doors.

The newsroom’s alert becomes effective through this work of association. The message has authors, but its capacity to inform and coordinate also depends on instruments, formats, connections, routines and responses.

Think of a performance

A singer’s performance draws on breath, training, a microphone, amplification, electrical supply, room acoustics and an audience. Change the microphone position and the sound changes. Lose the power supply and the performance must reorganise. ANT brings this attention to participation and dependence into social inquiry.

Three questions to carry through the chapter

• Composition: who and what make a particular action possible?
• Transformation: how do their connections redirect interests, meanings and conduct?
• Durability: which arrangements make the result repeatable, authoritative or difficult to change?

In my lecture, I introduce actors, heterogeneous networks, translation, black boxes and irreversibility. These concepts give us a route into the practical work through which media acquire reach and force.

Follow the work through which a collective is assembled.

ANT developed through science and technology studies, especially during the 1980s. Bruno Latour, Michel Callon and John Law examined how scientific claims, technical projects and organisations emerge through arrangements of people, materials, instruments and texts.

Laboratory studies offered a productive starting point. A finding takes shape through experiments, samples, skilled handling, instruments, calculations, publications and criticism. The same attention can illuminate a newsroom, a film production, a communications campaign or a platform.

A material-semiotic approach

“Material-semiotic” brings material arrangements and meaning-making into the same investigation. A flood gauge measures under particular physical conditions; a warning threshold classifies its reading; a red marker makes that classification visible. Each participates in what the alert can mean and do.

A collective names an association of humans and nonhumans. An assemblage draws attention to their coming together. These terms direct the researcher towards the relations that compose an entity and sustain its capacities.

The social as something to investigate

ANT asks how people and things become associated. An explanation such as “the organisation decided” becomes the beginning of an inquiry into meetings, permissions, documents, software and people authorised to decide.

This approach combines theoretical commitments with a method of investigation. Different ANT studies develop different emphases, and later work explores multiplicity, care, exclusion and the politics of research itself. The field has grown through internal debate as well as application.

Read alongside Latour (2005), Law (1992) and Law and Hassard (1999).

An actor makes a difference to what happens.

An actor is an entity whose participation changes a course of action. The term actant, drawn from semiotics, helps keep open the range of entities that can participate. A person, a file format, a microorganism, a measuring instrument or a rule embodied in a form may enter an investigation.

Agency is examined through effects in relations. A locked field in a publishing system can prevent a reporter from submitting a story. Its effect depends on code, permissions, organisational routines and the reporter’s attempts to proceed.

Distributed agency

A journalist’s capacity to publish emerges through a working arrangement: professional skill, access credentials, sources, equipment, deadlines and editorial approval. Action is distributed across these relations, while the participants contribute in different ways.

Generalised symmetry

Generalised symmetry asks the researcher to investigate human and nonhuman participation without deciding in advance that one category provides the entire explanation. Begin with the same willingness to follow their relations, then establish the differences that the evidence reveals.

Use a practical test: what changes when a participant is removed, replaced, bypassed or fails? A missing time stamp may alter a screenshot’s interpretation. A changed editor may alter the verification process. These comparisons sharpen the account of agency.

Nonhuman participants include living organisms as well as objects and technical systems. Their relevance comes from demonstrated participation in the event under study. Naming an entity as an actant creates a research question about its contribution.

Trace the connections, then examine what changes along them.

A network in ANT is an account of associations through which something happens. Follow a report from observation to record, from record to headline, and from headline to public use. At each connection, ask what is carried forward and what is transformed.

An actor can also be investigated as a network. “The newsroom” may function as one participant in a study of public communication. Opening it reveals editors, contracts, publishing tools, source relations and working routines. The level of detail follows the research question.

Intermediaries and mediators

An intermediary transports something in a way that can be treated as unchanged for the purpose of the account. A mediator transforms, translates, redirects or modifies what passes through it.

Suppose a file transfer preserves the approved bulletin exactly. The researcher may treat it as an intermediary in that episode. If a messaging service compresses an image until the location becomes unreadable, that service becomes a mediator of the alert’s use.

The same entity can function differently in different relations. A spreadsheet may preserve entered numbers during storage, then mediate their significance when a formula classifies some readings as dangerous. Follow the operation that matters.

Follow the actors

Actors often point towards the next connection: a reporter refers to a verification sheet; the sheet names a source; the source describes a calibration routine. Follow these leads while comparing accounts with observations and documents. A participant’s explanation is valuable evidence that can itself be investigated.

A shared project takes shape as interests and roles are reworked.

Translation describes the processes through which participants, interests and capacities are aligned and transformed. A flood-alert project may connect a newsroom’s demand for verification, residents’ need for usable directions, volunteers’ local knowledge and a funder’s demand for measurable reach.

The resulting arrangement changes the participants’ roles. A volunteer becomes an authorised translator; a gauge becomes an accepted source of warning evidence; a newsroom becomes a coordinator of information. These identities emerge through the work of alignment.

Callon’s study of scallops and fishermen

Callon’s account of scallop cultivation in St Brieuc Bay follows researchers, fishermen, scientific colleagues and scallops. It examines attempts to organise their interests and establish representatives. Resistance and failure reveal the fragility of the proposed alignment.

From this study, Callon distinguishes four moments of translation. Each directs attention to a different question about the formation of a collective.

The obligatory passage point

An obligatory passage point is a route that a translation project seeks to make indispensable for reaching participants’ goals. If the newsroom defines its verification desk as the required route for an authorised alert, it is attempting to establish such a passage point.

Its success depends on others’ participation. Residents may rely on another source; a department may withhold data; volunteers may create a parallel channel. Trace how indispensability is established and how it is challenged.

Spokespersons and representation

Mobilisation raises a demanding question: how does a spokesperson become entitled to represent others? A community leader’s statement, a survey summary and an instrument reading each represent something beyond themselves, through different procedures.

Check how the claim was assembled. Who was consulted? Which measurements were taken? Which people or conditions disappear when the result is condensed? A public challenge may reopen an apparently settled act of representation.

Foundational reading: Callon (1986). For a fuller media application, return to the cases in Section 12.

Documents and devices carry action across places and times.

An inscription gives an activity, observation or decision a durable and transportable form. A graph records measurements; a form records a resident’s report; a time stamp locates an alert in time. Such traces can be copied, compared, stored and used elsewhere.

Every inscription selects. A flood-warning form may record location, depth and time while leaving little space for uncertainty or access needs. Its categories shape what later appears in the dataset and what the editor can readily compare.

From experience to a usable record

Suppose a resident describes water rising “very quickly”. A reporter enters a numerical estimate in a required field. The description has crossed formats and acquired a new appearance of precision. Examine how the estimate was obtained and which qualifications travel with it.

Technical scripts

Madeleine Akrich’s account of scripts directs attention to the users and actions imagined within a device. A reporting application may anticipate one language, one personal phone and continuous connectivity. Actual use reveals how well these expectations fit people’s circumstances.

Users can follow, adapt, bypass or contest a script. A reporter may record a voice note offline and later ask a colleague to enter it. This workaround becomes part of the network that makes the application usable.

Delegation changes the work

Latour’s analysis of mundane artefacts shows how prescriptions can be carried by things. A door-closer helps maintain a closed door. In media work, a publication rule can move into software validation. That move changes who encounters the rule, how exceptions are handled and who must repair failures.

Compare a rule spoken at a meeting, printed in a manual and encoded in a form. Each has a different reach, vulnerability and demand for maintenance. This comparison reveals how organisation acquires material support.

Explore Akrich (1992), Latour (1992) and Latour (1987).

A working arrangement can come to appear as a single thing.

Black boxing occurs when the internal work of an arrangement recedes from attention and users deal mainly with its inputs and outputs. A phone becomes “the phone”; an editorial workflow becomes “the approval process”; a successful alert service becomes “the system”.

This simplification helps people act. A reporter can use a functioning recorder without inspecting every component. A newsroom can rely on a standard procedure while directing attention towards the next story.

Punctualisation

Law uses punctualisation for the process through which a heterogeneous network is treated as a single actor or resource. Black boxing and punctualisation draw attention to related simplifications: the many relations within an entity become available to others through an apparent unity.

Reliability requires work

Maintenance sustains a network’s apparent simplicity. Someone renews certificates, charges batteries, checks measurements, updates contacts and trains new staff. An account of successful communication becomes stronger when it includes this often overlooked work.

Star and Ruhleder’s study of infrastructure draws attention to relationships between systems, communities and access. What runs smoothly for an experienced user may demand extensive learning and support from someone entering the arrangement.

Reopening a settled claim

A trusted warning threshold may become a black box when staff routinely accept its classifications. An unexpected result can lead them back to calibration, categories and assumptions. The claim’s history becomes available for inquiry again.

For media research, follow both routine and disturbance. The smooth publication and the failed publication reveal different parts of the same organisation.

Read Law (1992) and Star and Ruhleder (1996).

An arrangement becomes durable when later action builds upon it.

Networks acquire stability through repeated use, compatible components, accepted standards, institutional support and continued maintenance. Each successful use may encourage further investment and make the arrangement more dependable for those involved.

Suppose several organisations adopt the newsroom’s warning categories. Training manuals use them, software fields require them, volunteers learn them and archived reports become searchable through them. The categories now have multiple supports.

Irreversibility as a matter of degree

Callon’s discussion of irreversibility examines how difficult it becomes to return to an earlier arrangement and how strongly an established translation shapes later ones. Replacing a category system may require new training, converted records, revised software and fresh agreements.

The difficulty lies in accumulated relations. A small change on a screen can carry a large organisational cost because so much work has come to depend on that field.

Resistance is evidence

A gauge gives unexpected readings. A reporter refuses a misleading category. A community organisation rejects a spokesperson. A user bypasses the prescribed application. Each episode reveals a limit to the proposed alignment.

Record how the project responds. It may modify roles, add resources, tighten controls, negotiate a compromise or abandon an element. These responses show the practical work required to maintain a collective.

Compare the paths that remained open

A useful historical inquiry follows alternatives. Which formats were considered? Why was one supplier selected? Who argued for another route? Which later decisions made a switch harder?

This sequence makes durability visible as an achievement with a history. It also identifies opportunities for intervention: an open export format, a parallel access channel or a revised representation procedure may loosen a dependency.

See Callon (1990). The reading list links the journal version of this discussion.

Trace the arrangements that allow an actor to act at a distance.

ANT investigates power through the associations that make commands, standards and decisions effective. A central office can coordinate distant teams when budgets, records, permissions, devices and staff carry its decisions into practice.

Scale becomes a research question. A national media organisation reaches across places through bureaux, schedules, contracts, distribution systems and accepted formats. Its size takes practical form in these arrangements.

From local records to central decisions

A centre of calculation gathers inscriptions from dispersed locations and makes them comparable. A newsroom may assemble reports on a dashboard, compare locations and allocate reporting teams. The power of the centre depends on records travelling into it and instructions travelling out.

Asymmetry has material supports

One participant may control an obligatory passage point, own essential equipment, set a classification or refuse a contract. These advantages can become durable when supported by many relations.

Analytical symmetry creates room to investigate these differences. A student contributor and a platform owner may both participate in a publishing network, while possessing very different rights, resources and opportunities to change it.

Organisations carry histories

Budgets, archives, property records and professional credentials bring earlier decisions into present action. Follow those records when they explain the current distribution of capacity. A present interface may carry a long history of procurement, labour and exclusion.

The researcher can therefore connect a small publishing episode to wider institutions by tracing the links that make those institutions effective. An account gains explanatory force when it shows how a decision reaches its destination, where it encounters resistance and what makes compliance possible.

A media product travels through a changing arrangement of participants.

ANT offers a way to connect media production, circulation and use. Begin with a bounded event, then follow the participants and transformations that carry it across these activities.

Plesner’s study treats communication technologies as participants in changing newswork. Domingo, Masip and Costera Meijer develop an approach to tracing digital news across production, circulation and use. These applications encourage close attention to the practices through which media boundaries are made.

Algorithms in practice

A ranking result can be studied through the arrangements that produce and use it: objectives, data categories, code, commercial agreements, moderation work and user responses. The investigation becomes specific when it follows a particular ranking decision or change.

Where access to a system is limited, make the evidence boundary explicit. A creator’s interpretation of falling reach, a visible interface change and an internal ranking rule are different kinds of evidence. A strong account distinguishes them and examines how participants act upon each.

Classification distributes visibility

Bowker and Star show how standards and categories organise lives and work. In a media setting, a language category or content label can affect discoverability, moderation and payment. Trace who designed the category, how it is applied and which cases strain its boundaries.

Audience participation extends the network

A viewer forwards a clip, adds a caption, translates a sentence or uses a video in a classroom. These practices can redirect a media object and connect it with new participants. A screenshot may carry a message while losing its source link and update history.

This approach complements questions about ownership, discourse and interpretation. ANT can show how a contractual advantage enters a workflow, how a category enters an interface, or how a preferred meaning changes during circulation. Define the contribution of each approach when combining them.

Media readings: Plesner (2009), Domingo and colleagues (2015) and Couldry (2008).

Keep the formation of the explanation open to investigation.

Latour’s Reassembling the Social organises inquiry around five sources of uncertainty. These encourage the researcher to examine how groups, actions, objects, facts and accounts take shape in practice.

Knowledge is made through demanding procedures

A claim gains strength through testing, measurement, criticism and reliable connections with the phenomena it describes. Study the instruments, procedures and disputes through which that strength is established.

In this approach, construction points towards the work of producing and sustaining knowledge. A carefully calibrated reading and an unsupported assertion can have very different evidential standing. The investigation follows how that difference is established and tested.

Strengths of the approach

• It makes overlooked participants visible. Forms, cables, maintenance workers and data categories enter the account when they change the outcome.
• It links meaning with material practice. A warning’s wording can be examined alongside the tools and routines that give it force.
• It explains stability through work. Repeated success, breakdown and repair reveal how an arrangement survives.
• It follows changing boundaries. Producers, audiences, tools and organisations can acquire new roles during an event.

Questions that sharpen an ANT study

Power and exclusion: a study centred on a successful project can give its organisers too much control over the account. Seek participants who refuse, remain outside or bear its costs. Star’s attention to marginal experience helps bring these consequences into view.

History and structural inequality: detailed tracing needs a reasoned account of where to begin and how far to follow a relation. Include ownership, caste, class, gender or colonial histories where their effects enter access, authority, resources and classification. Archival work can reveal links that present observation alone leaves obscure.

Meaning and public consequences: media operate through representation, interpretation and claims to speak for society. Couldry asks how ANT connects with media theory and where its reach requires further discussion. Examine both the machinery of circulation and the public significance of what circulates.

Ethical judgement: mapping contributions opens a route to identifying responsibility. Ask who chose the design, controlled the resources, could foresee a consequence and could change the arrangement. A statement that “the system did it” should lead to further investigation.

Multiplicity and research practice

Mol’s work examines how an object can be enacted differently across practices and made to cohere through coordination. For media, this suggests asking how an “audience” becomes a collection of viewers in one practice, a metric in another and a public in another.

Law’s After Method adds a reflexive question: how do our categories, observations and reports help organise the realities we describe? Keep a record of analytical decisions, absent evidence and alternative readings. This makes the research account open to scrutiny.

For these debates, consult Star (1990), Couldry (2008), Mol (2002), Law (2004) and the contrasting political questions in Winner (1980).

Use the concepts to reconstruct a concrete media event.

Each case begins with a communication problem. Identify the participants, trace at least two transformations, and examine one claim to represent a larger collective before opening the analysis.

Case 1 · The alert arrives without its time

Suppose a newsroom issues a flood alert at 7 a.m. A community administrator forwards a screenshot after cropping the page header. At noon, the water level has changed. Residents continue sharing the screenshot because its image loads quickly on weak connections. A volunteer adds an update, but the earlier screenshot keeps circulating.

Your task: explain how an accurate morning message can acquire a different practical effect. Identify a mediator, a lost inscription and a possible intervention.

Case 2 · A dashboard becomes the voice of the audience

Suppose a digital publication adopts a dashboard that ranks stories by visits. Editors use the ranking at daily meetings. Reporters begin pitching topics that resemble earlier winners. The system counts repeat visits, while some readers access stories through shared screenshots. A community desk argues that its work serves readers poorly represented in the dashboard.

Your task: identify the inscription, spokesperson and emerging obligatory passage point. Propose evidence that would test the dashboard’s claim to represent audience value.

Case 3 · An automated subtitle changes a public statement

Suppose a university media team uploads a Bengali interview with English subtitles generated by software. Background noise affects a sentence. A reviewer checks the video quickly, and a subtitle changes the speaker’s position on a campus issue. Viewers quote the subtitle. The speaker requests a correction; the team changes the uploaded file, while copied clips retain the earlier wording.

Your task: locate distributed agency and human accountability. Explain why a corrected original may leave part of the problem unresolved.

Listen for the shift from fixed entities to working associations.

Watch my lecture on Actor–Network Theory, then return to the flood-alert example. Listen for the connections among actors, heterogeneous collectives, translation, black boxes and the production of apparent scale.

Watch my Actor–Network Theory lecture on YouTube ↗

A three-pass viewing exercise

1. First pass: write a two-sentence explanation of why nonhuman participation matters.
2. Second pass: connect translation, enrolment, black boxing and durability to one media practice you know.
3. Third pass: select one claim about power or organisation and identify the evidence needed to investigate it.

Finish by drawing one association with a verb: a field requires a location; a time stamp locates an alert; a volunteer rewords instructions. These verbs turn an inventory into an account of action.

Ten questions to test conceptual precision.

Choose an answer before opening its explanation. Use an example to explain why the selected option fits.

Build an account that another researcher can follow.

Begin with a specific event: a story that changed during verification, a delayed upload, an inaccessible announcement, a contested audience metric or a correction that travelled unevenly. A bounded episode gives the investigation a clear point of entry.

Activity 1 · Open one communication black box

1. Select an ordinary practice, such as publishing a class notice or distributing a podcast.
2. Record the apparent input and output: what goes in, and what users expect to receive.
3. Identify at least eight participants, including people, material arrangements and documents.
4. Replace vague connections with verbs: authorises, converts, blocks, verifies, funds, classifies or forwards.
5. Follow one failure or workaround. Explain which hidden dependency it reveals.
6. Write 400–500 words linking three ANT concepts with specific evidence.

Activity 2 · Trace a representative claim

Choose a statement such as “the data show what viewers want” or “this committee represents students”. Follow the procedure through which people, observations or records become a single public statement. Compare the representative’s account with the experiences of at least two differently placed participants.

Produce a small map, a table of included and excluded contributions, and a 250-word account of one challenge to the claim. Identify how that challenge could change the arrangement.

A practical research workflow

Choose methods to match the relation

Observation shows work as it unfolds. Interviews reveal interpretations and leads. Documents preserve formal decisions and inscriptions. Interface walkthroughs show the user positions a system anticipates. Version histories and available logs help reconstruct changes. Compare these materials when they speak to the same event.

For controversy mapping, follow competing problem definitions, spokespersons, evidence and proposed solutions. Venturini’s work provides a route into this practice. A disagreement about automated subtitles, for example, can connect accessibility, editorial labour, language resources, cost and public accountability.

Set a defensible boundary

Follow connections that change the answer to the research question. Explain why some links receive close attention and others remain contextual. Mark where evidence ends, especially when internal software records or contractual documents are unavailable.

Research access also has an ethical dimension. Obtain appropriate consent for observation, protect sources and redact credentials or private messages from research outputs. Keep enough detail to support the account without exposing participants unnecessarily.

Examination prompts

• Short answer: define actant, mediator, inscription and black box with one media example each.
• Conceptual answer: explain Callon’s four moments of translation and the problem of representation.
• Applied essay: use ANT to investigate how a news alert becomes authoritative and usable.
• Critical essay: assess ANT’s contribution to media research, considering power, history, meaning and accountability.

A developed answer states the central approach, explains selected concepts, follows a concrete chain of associations and evaluates what the approach reveals. Let the example carry the argument from one concept to the next.

Keep the relations visible as you deepen the inquiry.

Use the glossary to recover the argument behind each term. Begin further reading with Latour, Law and Callon, then follow the media applications and critical extensions that match your question.

Glossary

Actor

An entity whose participation makes a difference to a course of action.

Actant

A term that keeps human and nonhuman participation open to investigation.

Agency

The capacity to make a difference, examined through relations and effects.

Distributed agency

Action emerging through the contributions of multiple connected participants.

Human participant

A person whose practices, decisions or capacities enter the account.

Nonhuman participant

A living entity, material object or other nonhuman element that affects the event.

Heterogeneity

The involvement of different kinds of participants, materials and capacities.

Material-semiotic

An approach that studies material arrangements and meaning-making through their relations.

Association

A connection through which participants become involved in a course of action.

Actor-network

An actor understood through the relations that compose and sustain its capacities.

Collective

An association composed of humans and nonhumans.

Assemblage

A coming together of participants whose relations and effects require investigation.

Generalised symmetry

An analytical commitment to investigating human and nonhuman contributions without assigning priority beforehand.

Intermediary

Something treated as transporting a contribution without consequential transformation in the account.

Mediator

A participant that transforms, redirects or modifies what passes through it.

Translation

Processes that align and transform participants, identities, interests and capacities.

Problematisation

Defining a situation, its participants and a proposed route through the problem.

Interessement

Attempts to secure participants within proposed relations and roles.

Enrolment

The negotiation and establishment of interconnected roles within an arrangement.

Mobilisation

Processes through which representatives become able to speak or act for a collective.

Obligatory passage point

A route that a project seeks to make indispensable to participants’ goals.

Spokesperson

A participant authorised or accepted as representing others.

Representation

A relation through which someone or something stands for a wider collective or condition.

Inscription

A durable, transportable trace such as a graph, document, recording or dataset.

Inscription device

An arrangement that produces a trace through which something becomes recordable and comparable.

Immutable mobile

A representation that travels while retaining enough stable form to support comparison or combination.

Script

Expectations about users and conduct built into an artefact or arrangement.

Delegation

The transfer of part of a task or prescription to another participant or arrangement.

Black box

An arrangement whose internal work recedes behind apparent input-output reliability.

Black boxing

The process through which that internal complexity becomes taken for granted.

Punctualisation

The treatment of a heterogeneous network as a single actor or resource.

Stabilisation

The work through which relations become sufficiently durable and repeatable.

Maintenance

Activities that sustain the functioning and reliability of an arrangement.

Breakdown

A failure that can expose dependencies and prompt reorganisation.

Resistance

Conduct or material effects that interrupt a proposed alignment.

Irreversibility

The difficulty of undoing an established translation and its influence on later arrangements.

Standard

An accepted specification that coordinates practices and supports compatibility or comparison.

Centre of calculation

A site that gathers and combines inscriptions to compare conditions and coordinate action.

Action at a distance

The capacity to affect activity elsewhere through maintained connections.

Controversy

An unsettled issue in which participants, evidence and proposed relations become visible through disagreement.

Multiplicity

The enactment of an object differently across practices, with questions about coordination and coherence.

Reflexivity

Attention to how research choices and accounts participate in arranging what is studied.

References and further reading

1. Latour, B. (2005). Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford University Press. Read / view source ↗A sustained introduction to associations, agency, mediation and the five uncertainties.
2. Law, J. (1992). Notes on the theory of the actor-network: Ordering, strategy, and heterogeneity. Systems Practice, 5, 379–393. Read / view source ↗A concise account of heterogeneous networks, ordering, power and punctualisation.
3. Callon, M. (1986). Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St Brieuc Bay. In J. Law (Ed.), Power, Action and Belief: A New Sociology of Knowledge? (pp. 196–223). Routledge & Kegan Paul. Read / view source ↗The foundational study of problematisation, interessement, enrolment and mobilisation.
4. Akrich, M. (1992). The de-scription of technical objects. In W. E. Bijker & J. Law (Eds.), Shaping Technology/Building Society: Studies in Sociotechnical Change (pp. 205–224). MIT Press. Read / view source ↗Develops the analysis of technical scripts and anticipated users.
5. Latour, B. (1992). Where are the missing masses? The sociology of a few mundane artifacts. In W. E. Bijker & J. Law (Eds.), Shaping Technology/Building Society: Studies in Sociotechnical Change. MIT Press. Read / view source ↗Examines how mundane artefacts participate in prescriptions and collective life.
6. Latour, B. (1987). Science in Action: How to Follow Scientists and Engineers through Society. Harvard University Press. Read / view source ↗Follow knowledge-making, inscriptions, laboratories and extended relations.
7. Latour, B., & Woolgar, S. (1986). Laboratory Life: The Construction of Scientific Facts (2nd ed.). Princeton University Press. Read / view source ↗An influential ethnographic investigation of scientific work. The first edition appeared in 1979.
8. Callon, M. (1990). Techno-economic networks and irreversibility. The Sociological Review, 38(1_suppl), 132–161. Read / view source ↗Explores coordination, translation and the durability of network arrangements.
9. Law, J., & Hassard, J. (Eds.). (1999). Actor Network Theory and After. Blackwell / The Sociological Review. Read / view source ↗A collection that develops debates about complexity, subjectivity, materiality and politics.
10. Latour, B. (1999). Pandora’s Hope: Essays on the Reality of Science Studies. Harvard University Press. Read / view source ↗Extends the investigation of science, reference, reality and mediation.
11. Star, S. L. (1990). Power, technology and the phenomenology of conventions: On being allergic to onions. The Sociological Review, 38(1_suppl), 26–56. Read / view source ↗Brings marginal experience and the consequences of conventions into the analysis.
12. Star, S. L., & Ruhleder, K. (1996). Steps toward an ecology of infrastructure: Design and access for large information spaces. Information Systems Research, 7(1), 111–134. Read / view source ↗Connects infrastructure, access, learning and communities of practice.
13. Bowker, G. C., & Star, S. L. (1999). Sorting Things Out: Classification and Its Consequences. MIT Press. Read / view source ↗Shows how categories and standards organise information and distribute consequences.
14. Mol, A. (2002). The Body Multiple: Ontology in Medical Practice. Duke University Press. Read / view source ↗Examines how objects are enacted differently across practices and made to cohere.
15. Law, J. (2004). After Method: Mess in Social Science Research. Routledge. Read / view source ↗Develops questions about methods, multiplicity and the politics of producing research accounts.
16. Couldry, N. (2008). Actor network theory and media: Do they connect and on what terms? In A. Hepp, F. Krotz, S. Moores, & C. Winter (Eds.), Connectivity, Networks and Flows: Conceptualizing Contemporary Communications (pp. 93–110). Hampton Press. Read / view source ↗Assesses connections and limits between ANT and media theory.
17. Plesner, U. (2009). An actor-network perspective on changing work practices: Communication technologies as actants in newswork. Journalism, 10(5), 604–626. Read / view source ↗A direct application to the relations among technologies and journalistic work.
18. Domingo, D., Masip, P., & Costera Meijer, I. (2015). Tracing digital news networks: Towards an integrated framework of the dynamics of news production, circulation and use. Digital Journalism, 3(1), 53–67. Read / view source ↗Develops ANT-informed investigation across production, circulation and use.
19. Venturini, T. (2010). Diving in magma: How to explore controversies with actor-network theory. Public Understanding of Science, 19(3), 258–273. Read / view source ↗Introduces the exploration of controversies as a research and teaching practice.
20. Winner, L. (1980). Do artifacts have politics? Daedalus, 109(1), 121–136. Read / view source ↗A complementary political inquiry into technology, authority and forms of social order.

Six ideas to retain

1. Begin with associations. Show how a course of action is assembled.
2. Examine heterogeneous participation. Establish what people, organisms, objects and records contribute.
3. Follow transformations. Identify mediators, translations and changes in representation.
4. Open apparent unities. Trace the dependencies behind a person’s capacity, a device or an organisation.
5. Explain durability. Include standards, maintenance, resistance and the costs of change.
6. Make the account inspectable. Connect claims with evidence, boundaries and consequences.