Giddens - Structuration and the Duality of Structure

Anthony Giddens (b. 1938) produced the other major attempt to dissolve the structure–agency dichotomy, and it takes the opposite route from Bourdieu's.

Bourdieu puts structure inside the actor, as durable disposition. Giddens puts it inside the act — arguing that structure has no existence at all outside the moments in which it is produced, and that every action simultaneously draws on structure and remakes it.

Say a sentence.

Say: the meeting has been moved to Thursday.

Now ask what made that possible.

English grammar. A vocabulary. Rules of word order. Conventions about tense. None of which you consulted, and all of which constrained you absolutely — say the words in a different order and you produce nothing.

So: the language constrained you.

And now ask where the language is.

Not in a book — dictionaries and grammars are descriptions written after the fact, and they change when usage does. Not in any one head — nobody knows all of English, and everyone's version differs. Not in an institution — no body controls it, and the institutions that have tried have failed.

English exists only in the instances of its use. In sentences, spoken and written, by people who mostly could not state the rules they are following.

And here is Giddens's move. Your sentence did not merely obey English. It reproduced it. Every competent utterance is one more instance confirming that these are the rules, and every slight variation that is understood and copied is a small alteration to them.

Structure constrained the act, and the act reproduced the structure, and there is no moment at which the two are separate.

"Structure is both the medium and the outcome of the practices it recursively organises."

That sentence is structuration theory, and everything else in this lesson is its elaboration.

Giddens's problem is the one Part 5 keeps returning to, and his statement of the impasse is exact.

The orthodox consensus — functionalism and structuralism together — treated structure as external and constraining: something out there that acts on people, whose properties can be studied independently of what anyone does.

The interpretive alternatives — interactionism, phenomenology, ethnomethodology — treated social life as produced in interaction, with structure appearing as an unhelpful reification.

Both, Giddens argues, share a mistake: they assume structure and agency are two different things that must be related. Once you accept that, you must either subordinate one to the other or produce an unstable compromise.

His alternative is to deny the dichotomy at the outset. Structure is not a thing that constrains action from outside. It is a property of action — and specifically of the repetition of action across time and space.

Which requires rebuilding the concept of structure itself.

The duality of structure

Giddens's redefinition: structure consists of rules and resources which exist only as memory traces in the minds of knowledgeable actors, and are instantiated in action.

Rules — generalisable procedures applied in the enactment of social practices. Not written regulations but the tacit know-how of how to go on: how to queue, how to greet, how to make a promise, how to complain.

Resources — the media through which power is exercised, of two kinds: Allocative — control over material objects and goods. Authoritative — control over persons and their activities.

And the crucial claim, which is the theory's whole content:

"The duality of structure: structure is both medium and outcome of the reproduction of practices. Structure enters simultaneously into the constitution of the agent and social practices, and 'exists' in the generating moments of this constitution."

Duality, not dualism. A dualism is two things requiring a relationship. A duality is two aspects of a single process.

Structure / system / structuration — Giddens uses these precisely, and confusing them makes the theory incoherent.

Structure — rules and resources, existing out of time and space, as memory traces and instantiations. Structure has only a "virtual existence."

System — the actual, observable, patterned relations between actors, reproduced across time and space. Systems have structural properties; they are not structures. A firm, a family, a state is a system.

Structuration — the process by which systems are produced and reproduced through the application of structural properties in interaction. The verb form is the theory.

And Giddens's most quoted formulation of what this does to constraint:

"Structure is not to be equated with constraint but is always both constraining and enabling."

The language that limits what you can say is the same language that lets you say anything at all. Rules do not merely restrict; they make action possible. A game without rules is not freedom; it is not a game.

The knowledgeable agent

Giddens's actor is far more competent than most theories allow, and this is a substantive claim rather than a compliment.

Actors are knowledgeable. They know a great deal about the conditions of their own action — far more than they can articulate.

Three levels of consciousness:

Discursive consciousness — what actors can say about their action: reasons they can state if asked.

Practical consciousness — what actors know how to do without being able to say. This is the largest and most important category, and it is Giddens's debt to Garfinkel (2.4.3): the tacit competence required to be a member of a society. You know how close to stand, how to take a turn in conversation, what tone is appropriate. Ask why, and you will produce a rationalisation.

The unconscious — motives not accessible to the actor, and Giddens takes the psychoanalytic tradition seriously here, particularly on ontological security (below).

And the stratification model of action: actors engage in reflexive monitoring of their conduct and its settings, rationalisation (maintaining a continuing theoretical understanding of the grounds of their activity), and motivation — which is much less continuously involved than the first two, since most action is not directly motivated but flows from routine.

Unintended consequences and unacknowledged conditions, which is where structuration explains reproduction.

Action occurs within unacknowledged conditions and produces unintended consequences, which then become the unacknowledged conditions of further action. The loop is how systems reproduce themselves without anyone intending it.

Giddens's own example is Marx's: in speaking English, nobody intends to reproduce the English language, and everybody does.

And a harder one: a person from a poor background who leaves school early to earn, acting reasonably on their circumstances, contributes to the aggregate pattern of educational inequality. They intended none of it, and it is the outcome of their action.

This is the structuration answer to Bourdieu's Karim (5.8.1) — and note the difference. Bourdieu explains the choice by the disposition; Giddens explains the reproduction by the aggregation of knowledgeable but partially sighted action. Both are accounts of the same phenomenon and they are not the same account.

Time-space, routine and ontological security

Giddens's most distinctive substantive contributions, and the ones most underused.

Time-space distanciation — the "stretching" of social systems across time and space. In a face-to-face society, social relations are bounded by co-presence. Modern institutions extend relations across enormous distances of time and space, through writing, money, clocks, records, transport and telecommunication.

Disembedding — the "lifting out" of social relations from local contexts and their restructuring across indefinite spans of time-space, through two mechanisms: Symbolic tokens — media of exchange passing between parties without regard to their characteristics; money being the type case (4.4.4). Expert systems — systems of technical accomplishment organising large areas of the material and social environment, in which we place trust without understanding. You board an aircraft trusting an entire apparatus of engineering, maintenance, air traffic control and regulation, none of which you can assess.

Routinisation and ontological security. Drawing on Erikson and on Garfinkel's breaching experiments (2.4.3), Giddens argues that the routine character of daily life is not trivial. It provides ontological security — the confidence that the natural and social worlds are as they appear, which is the psychological foundation of being able to act at all.

This is why breaching experiments produce such disproportionate distress, and why disruption of routine — displacement, imprisonment, sudden unemployment, bereavement — produces anxiety out of proportion to its practical consequences. The routine is doing psychological work nobody notices until it stops.

Return to Blumer's queue (5.3.1) and run Giddens's apparatus, because the comparison shows what each theory adds.

The structure: rules about first-come-first-served, spatial arrangement, appropriate distance, what constitutes joining. Nowhere written. Existing as memory traces in the minds of everyone present.

The instantiation: twelve people apply these rules and thereby produce a queue. Nobody consults them; this is practical consciousness.

And the duality: the rules enabled the queue (without them, twelve people at a bus stop is a crowd, and boarding is a scramble) and were reproduced by it (every orderly boarding confirms that this is how it is done).

Now the additions Giddens makes that Blumer does not.

Resources. Queues have authoritative resources in play: who can enforce, who can challenge, whose challenge carries weight. An elderly person's challenge to a queue-jumper carries differently from a child's. Power is in the queue, not outside it.

Time-space. The queue exists because of a timetable, a route, a ticketing system, and a transport authority — structures stretching far beyond the pavement, which the twelve people instantiate locally without seeing.

Unintended consequences. Nobody at the stop intends to sustain a national norm of queueing. Their behaviour does.

And the reproduction of a system. The queue is a system — patterned relations reproduced across time. Tomorrow's queue will have different people and the same structural properties.

What Bourdieu would add, and it is not trivial: that some people are more comfortable in queues than others, that the confidence to challenge a jumper is unequally distributed, and that this distribution tracks position. Giddens's actors are all similarly knowledgeable; Bourdieu's are differently disposed. That is the sharpest difference between them, and Bourdieu's version has more purchase on inequality.

Go deeper

The criticisms, which are substantial

Structuration attracted more sustained criticism than almost any theory of its period, and much of it landed.

It is not researchable. The most damaging charge. Nicos Mouzelis and others argue that by defining structure as virtual and existing only in instantiation, Giddens makes it impossible to study structural properties independently of action — which is what most sociological research does. Empirical applications of structuration are notably thin, and where they exist they frequently use it as a vocabulary rather than as a research design.

Giddens's own reply is that structuration is not a research programme but a set of "sensitising devices" (5.3.1's Blumer, invoked). Critics find this an evasion: a theory that guides nothing in particular has not solved the problem it claimed to solve.

It dissolves structure too far. Margaret Archer's critique is the most important. Her morphogenetic approach argues that structure and agency have different temporalities — structures pre-date the actions that reproduce or transform them, and post-date them. Analytical dualism is therefore necessary: they must be held apart analytically in order to study their interplay over time, even if they are inseparable in reality.

Archer's charge is central conflation: Giddens has fused structure and agency so completely that neither can be examined for its effect on the other. Her three-phase model — structural conditioning → social interaction → structural elaboration (or reproduction) — restores the sequence and makes research possible.

This criticism is widely regarded as decisive, and Archer's framework has had more empirical uptake than Giddens's.

It underplays constraint and power. Actors in structuration are knowledgeable, capable and able to "act otherwise" — Giddens's definition of agency is the capacity to have done otherwise. Critics note that some actors can act otherwise a great deal more than others, and that a framework in which everyone is a knowledgeable agent has difficulty registering the difference between a chief executive and a cleaner. Bourdieu's differently-disposed actors handle this better.

And it is eclectic to the point of incoherence, assembling Wittgenstein, Garfinkel, Goffman, Heidegger, Parsons, Marx and time-geography into a synthesis some readers find integrative and others find a list.

Giddens's second career

An unusual feature of this figure, and it belongs in the record.

From the early 1990s Giddens turned from social theory to the analysis of late modernity — The Consequences of Modernity (1990), Modernity and Self-Identity (1991), The Transformation of Intimacy (1992) — and then to politics.

His late-modern arguments (developed further in 5.9): the reflexive project of the self, in which identity becomes something to be constructed and continuously revised rather than inherited; pure relationships, entered for their own sake and sustained only while they satisfy both parties; manufactured risk, produced by human intervention rather than by nature; and detraditionalisation.

And then: The Third Way (1998), Director of the London School of Economics (1997–2003), adviser to Tony Blair, and a life peerage. He became the theorist of a political project, which is a rare trajectory for a social theorist and an instructive one.

The relevance to this course: Giddens is the clearest modern case of the question Weber posed (4.3.9) and Du Bois answered differently (4.5.6). His critics — including many sympathetic to structuration — argue that the political engagement thinned the theory into slogans. His defence is close to Habermas's (5.4.2): that a theory of reflexive modernity implies that social analysis enters public argument as a matter of course.

What survives

The duality of structure as a corrective. Even critics accept the core insight: structure has no existence apart from the practices that instantiate it, and treating structures as things that act is a reification. This is now a background assumption in much of the discipline.

Practical consciousness. The recognition that most social competence is tacit, and that asking people why they did something produces rationalisation rather than cause, is standard — and it is a corrective to method as much as to theory.

Time-space distanciation and disembedding. These have been enormously productive in globalisation research (11.6) and in the sociology of trust and expertise. Expert systems in particular is one of the better concepts for analysing modern dependence.

Ontological security. Widely used in the sociology of housing, health, migration and disaster — and it is one of the few concepts connecting sociological and psychological levels without reducing either.

And the debate it provoked. Archer's morphogenesis, Mouzelis's critique, Stones's "strong structuration," and the critical realist tradition around Roy Bhaskar — which holds that structures are real, causally efficacious, and irreducible to the practices that reproduce them — are all responses to structuration, and several are more useful than the original.

"Structuration says structure and agency are the same thing." It says they are two aspects of one process — a duality, not an identity — and Archer's charge is that fusing them prevents studying either.

"Structure is just rules." Rules and resources, allocative and authoritative. The resources are where power enters, and they are frequently dropped from summaries.

"Structure only constrains." Giddens's central corrective: structure is "always both constraining and enabling." Language is the type case.

"Practical consciousness is the unconscious." It is what actors know how to do but cannot articulate — accessible in principle, and different from repressed material.

"Structuration is a research method." Giddens explicitly denied this and offered it as sensitising devices, which critics regard as an evasion of the researchability problem.

"Giddens solved the structure–agency problem." He reframed it influentially. Archer's analytical dualism is widely regarded as the more usable response, and the problem is not closed.

Why it matters

It states the corrective that stuck. Structures do not act. They have no existence apart from the practices that instantiate them, and any sentence in which "the system" or "society" does something is shorthand that must be unpackable into what people did. That discipline is now general, and Giddens is why.

It names the constraint/enablement duality. The recognition that the same rules which limit also make possible is a genuine correction to a discipline that had treated structure almost exclusively as restriction.

It supplies the best account of tacit competence. Practical consciousness explains why interviews produce rationalisations, why breaching experiments distress, and why most of what people know about their society cannot be stated — which has consequences for every method in Part 7.

And it demonstrates the cost of dissolving a distinction too completely. Giddens fused structure and agency so thoroughly that neither could be studied, and the field's response — Archer's analytical dualism, critical realism, "strong structuration" — was to put back enough separation to make research possible. That is a useful lesson about theory: a dichotomy can be dissolved into incoherence as easily as it can be maintained into sterility.

"Structure is both the medium and the outcome of the practices it recursively organises." Speaking a sentence draws on English and reproduces it, and there is no moment at which the two are separate.

Structure = rules and resources (allocative and authoritative), existing virtually as memory traces and instantiated in action. System = actual patterned relations reproduced across time and space. Structuration = the process by which systems are produced through the application of structural properties. Duality, not dualism.

Structure is "always both constraining and enabling."

The knowledgeable agent operates at three levels: discursive consciousness (what can be said), practical consciousness (tacit competence — the largest category, from Garfinkel), and the unconscious. Action involves reflexive monitoring, rationalisation and motivation, and reproduces systems through unintended consequences becoming unacknowledged conditions.

Time-space distanciation, disembedding through symbolic tokens (money) and expert systems (trust without understanding), and routine supplying ontological security — which explains why breaching produces distress disproportionate to its consequences.

The criticisms: it is not researchable (Mouzelis), and Giddens's "sensitising devices" reply is regarded as an evasion; Archer's charge of central conflation, with her morphogenetic alternative — analytical dualism, and the sequence structural conditioning → interaction → elaboration — which has had far more empirical uptake; and the underplaying of differential capacity to act otherwise, which Bourdieu handles better.

What survives: the anti-reification corrective, practical consciousness, time-space distanciation and expert systems, ontological security, and the debate it provoked — Archer, Mouzelis, Stones, and critical realism.

Duality of structure — structure as both medium and outcome of the practices it organises.

Rules and resources — the components of structure; resources are allocative (material) and authoritative (over persons).

Structure / system / structuration — virtual rules and resources; actual patterned relations; the process connecting them.

Constraining and enabling — the double character of structure.

Discursive consciousness — what actors can articulate about their action.

Practical consciousness — tacit competence that actors cannot articulate.

Reflexive monitoring — the continuous purposive attention actors give to their conduct and its settings.

Unacknowledged conditions / unintended consequences — the loop by which systems reproduce without intention.

Time-space distanciation — the stretching of social relations across time and space.

Disembedding — the lifting of relations out of local contexts, via symbolic tokens and expert systems.

Expert systems — technical systems in which we place trust without understanding.

Ontological security — the confidence in the continuity of the world, supplied by routine.

Analytical dualism — Archer's requirement that structure and agency be separated analytically to study their interplay.

Morphogenesis — Archer's three-phase model: structural conditioning, social interaction, structural elaboration.

Central conflation — Archer's charge that structuration fuses structure and agency beyond examination.

One — find a structure you reproduce. Pick a convention you follow without thinking — queueing, greeting, turn-taking, dress. Ask where it exists, and note that the only answer is: in the instances.

Two — test practical consciousness. Try to state the rules governing how close you stand to people. You will fail, and you will follow them perfectly. That gap is Giddens's largest category.

Three — look for constraint that enables. Take a rule you find restrictive and ask what it makes possible. Traffic law, grammar, professional standards, queueing. The enabling half is usually invisible.

Four — trace an expert system. List what you trusted today without being able to assess it: water supply, food safety, the aircraft, the bridge, the medication, the payment system. Then ask what your trust actually rests on.

Five — apply Archer's sequence. Take any change you know about in an organisation. Identify the prior structural conditions, the interaction, and what was elaborated or reproduced. The sequence makes research possible where the duality does not.