Structure vs Agency

The three master questions in Topic 1.2 were the discipline's problems. This topic is about its tensions — six arguments that never get settled, not because sociologists are careless, but because both sides of each one are holding on to something true.

You will meet all six for the rest of your life in this subject. Learn to argue every side calmly and you will already be ahead of most people who have read a great deal more than you.

We begin with the deepest of them. Every other tension is, in some way, a version of this one.

Meera is twenty-three. She is a nurse in a private hospital in a mid-sized Indian city, and she is good at it.

Ask her how she got there and she will tell you a story about choosing. She will say she always liked biology. She will say she wanted to help people. She will say her cousin was a nurse and made it sound interesting. She will say — and she will mean it — I chose this.

Now let us assemble that choice from the outside.

Meera is the first in her family to finish school, so an unfamiliar profession was a large risk and nursing was the one profession her parents had actually seen a woman do respectably. Medicine would have needed four more years of fees her family could not raise and a coaching course they had never heard of. Engineering was discussed and dropped: the college was in another city, and a daughter living alone was not something her father could explain to his brothers. Teaching was the other candidate; nursing paid more. Her school had one careers talk, given by a nursing college. The government had recently expanded nursing seats. And in her state, at that time, nursing was overwhelmingly done by women — so much so that a male classmate who considered it was mocked out of the idea.

Every one of those facts was in place before Meera formed a preference.

And yet: her cousin's daughter, one street away, in almost identical circumstances, refused the whole arrangement, fought her family for two years, and is now an accountant in Hyderabad.

So — did Meera choose, or was she placed?

Both answers are obviously right, and that is the problem.

Say "she was placed" and you have explained the pattern beautifully — why nursing, why then, why women, why not medicine — but you have turned a thoughtful person into a billiard ball, and you cannot explain the cousin at all.

Say "she chose" and you have honoured her as a person — but you have made it a coincidence that thousands of young women in the same state made the same "free choice" in the same decade, and you are quietly on your way to concluding that everyone gets roughly what they picked.

This is the structure–agency problem. It is not a puzzle waiting for someone clever to solve. It is a permanent tension that every serious theory in sociology has to take a position on — and how it positions itself tells you most of what you need to know about it.

The idea, in plain words

Two clarifications that will save you from the most common confusions.

First, structure is not "society over there" and agency is not "the individual". Structures are made of people acting. Agency is exercised by people who are made of their society. The two are not two separate substances in a tug of war; that image is exactly what the best theories try to escape.

Second, structure does not only constrain — it also enables. This is the single most useful thing to understand here, and the standard example is language. The rules of a language are not up to you: you did not invent the grammar, and you cannot simply decide that a word means something else. That is a constraint. But those same rules are the only reason you can say anything at all, including something no human has ever said. Structure is the piano, not the handcuffs. You cannot play any note you like; you can play music that has never existed.

Where the theories stand

Almost every school you will meet in Part 5 can be located by how far it leans.

Leaning towards structure

Émile Durkheim (4.2) insisted that social facts are external to individuals, general across a population, and coercive — they push back when you resist them. He wanted sociology to explain social things by other social things, not by individual psychology. On this view, asking Meera why she chose nursing is almost beside the point; you should be asking why the rate of women entering nursing in that state rose in that decade.

Talcott Parsons (5.1) built an enormous system in which individuals largely occupy roles that the social system needs filled.

Louis Althusser (5.6) went furthest of all. For him individuals are interpellated — hailed, addressed, called into position — by ideology, and what we experience as free choice is the feeling of occupying a slot that was already cut for us. He spoke of history as "a process without a subject." It is the most uncompromising structural position in the discipline, and it is worth understanding precisely because it makes the difficulty vivid.

Leaning towards agency

Max Weber (4.3) demanded verstehen — that we grasp the meaning an actor attaches to their own action. You cannot explain what Meera did without knowing what nursing meant to her.

Symbolic interactionism (5.3) goes further: society is not a thing that acts on people; it is what people are doing when they interpret each other and respond. Herbert Blumer's premise is that people act towards things on the basis of the meanings those things have for them, and that those meanings arise in interaction and get revised in it.

Rational choice and exchange theories (5.5) begin from individuals pursuing goals under constraints, and try to build the big picture upward from there.

The great syntheses

By the 1970s and 80s, most theorists concluded that both extremes were untenable and set out to dissolve the dichotomy. These attempts are among the highest achievements of modern theory, and you will study them properly in 5.10.

Anthony Giddens — structuration. His move is elegant: structure and agency are not two things but two aspects of one process, "the duality of structure." Structure is both the medium and the outcome of action. Language again: you can only speak because English exists; English only continues to exist because people keep speaking it. Every time you use a rule, you reproduce it — and if you use it slightly differently, you have altered it a little. Society is thus continuously being made by the very people it is making.

Pierre Bourdieu — habitus and field. Bourdieu's answer is quieter and, many think, deeper. Recall habitus from 1.1.3: the dispositions you acquired from your position — taste, posture, sense of what is "for people like us." Habitus is structure inside the person. So when Meera "chooses," she is choosing genuinely, with real deliberation, from a menu that her habitus has already pre-sorted into thinkable and unthinkable. Bourdieu's phrase for the result is that people develop "a feel for the game." Nobody forces the outcome; the outcome is nonetheless highly predictable. This is the best available account of why free choices aggregate into rigid patterns.

Margaret Archer — analytical dualism. Archer objected that Giddens fuses structure and agency so tightly that you can no longer study them separately. Her alternative: keep them analytically distinct and track them over time. Structures pre-date the people who confront them; people then reproduce or transform them; the result is the structure the next generation inherits. She calls this the morphogenetic cycle, and it is especially useful for historical work.

Norbert Elias — figurations. Elias disliked the whole vocabulary. He argued we should speak not of "individual and society" but of figurations — shifting webs of interdependent people, like a game in progress. No player controls the game; the game exists only in the playing.

Go deeper

Agency is not equally distributed

Here is the point most introductory treatments miss, and it is essential.

The structure–agency debate is often posed as if it were about human beings in general: are we free? But sociologically, the more revealing question is: who gets how much room to manoeuvre, and why?

A wealthy person facing a bad job market has options: wait, retrain, move, live off savings, use contacts. A poor person facing the same market has almost none of these. A man in many societies can walk home at night making decisions about his route; a woman in the same city is making a different calculation with fewer branches. A citizen can complain to an authority; an undocumented migrant cannot, and both know it.

So agency is itself distributed by structure. Amartya Sen's capability approach (8.8) is essentially an attempt to measure this: not what people are formally permitted to do, but what they are actually able to do. Once you see agency as a resource rather than a universal human attribute, the whole debate becomes sharper and more useful.

The moral stakes are enormous

This is not an abstract quarrel. Where you place the dial has consequences.

Push everything to agency and you get a politics of individual responsibility: the poor chose poorly, the unemployed should try harder, the failing school has bad pupils. Push everything to structure and you get either paternalism (people cannot be trusted to run their own lives) or fatalism (nothing can be done until the whole system changes), and you lose the ability to hold anyone to account for anything.

Marx, of all people, gave the most balanced formulation, and it is worth memorising:

"Men make their own history, but they do not make it as they please; they do not make it under self-selected circumstances, but under circumstances existing already."

Both halves. Always both halves. Part 13 returns to how to hold them together without collapsing into either.

Unintended consequences and the limits of intention

One more reason the dichotomy is too crude: even when people act with full agency and complete knowledge of their own aims, the aggregate result is usually something nobody intended.

Every parent in a city individually decides to coach their child harder to secure an advantage. Because everyone does it, no one gains an advantage, and childhood becomes measurably worse for all of them. Every commuter individually chooses the faster road; the road is now slow. Everyone acted freely; the outcome resembles a structure imposed from above.

Sociologists call this emergence — the appearance, at the collective level, of properties that no individual possesses or chose (see 1.3.2). It is one of the strongest arguments for the reality of structures: they are not mystical, they are what individual actions add up to, and they are nonetheless not reducible to anyone's decision.

"So it's 50–50." No. The mix is a genuinely empirical question and it varies enormously by case, by domain and by position. In some situations structure explains nearly everything; in others a single person's decision changed the outcome. Percentages are the wrong image; the right question is how they interlock in this case.

"Structure = society, agency = the individual." This is the framing all the good theories were built to escape. Structures exist only because people keep enacting them; people are themselves formed by structures. Giddens, Bourdieu, Archer and Elias are four different ways of saying so.

"Structure just means constraint." It also means capacity. Language, money, law, institutions and roles are all structures, and every one of them lets you do things that would be impossible without them.

"Agency means resistance." A common romantic error. Meera's careful, competent, entirely conformist decision was exercised agency too. Most agency is used to get by within an arrangement, not to overturn it — and reserving the word for rebellion makes ordinary people look like passive scenery.

"If structures shape us, no one is responsible." Two different questions: what explains an action, and who is answerable for it. Explanation can be full and responsibility can still hold. Confusing the two is how a real insight gets turned into an excuse.

Why it matters

This tension is the hinge on which the practical use of sociology turns.

In your own life, it is the difference between two bad stories — "everything is my fault" and "nothing is up to me" — and a third, more accurate one: here is the room I actually have, and here is what would have to change for it to be bigger. That third story is the only one you can act on.

In public argument, almost every dispute about welfare, crime, education, addiction, unemployment and inequality is a structure–agency dispute in disguise. Learning to spot the dial being moved is one of the most transferable skills the discipline offers.

In research, your position on this tension decides your method. If you lean structural you will count things — rates, distributions, flows (7.3, 7.4). If you lean towards agency you will listen to people and watch them closely (7.5). The best work does both, which is exactly why Part 7 refuses to let you specialise too early.

Structure is what is already in place before you act; agency is your capacity to interpret, choose and act otherwise.

Both are real. Explanations that use only one of them either turn people into billiard balls or turn patterns into coincidences.

Durkheim, Parsons and Althusser lean structural; Weber, interactionism and rational choice lean towards agency.

The great syntheses dissolve the opposition: Giddens (structure as both medium and outcome of action), Bourdieu (habitus — structure inside the person, pre-sorting the thinkable), Archer (analytical dualism over time), Elias (figurations).

Structure both constrains and enables — the piano, not the handcuffs.

Agency is unequally distributed: how much room a person has to manoeuvre is itself a product of their position.

Because of emergence, even wholly free actions aggregate into outcomes nobody chose.

Structure — the durable, patterned arrangements (institutions, rules, resources, distributions, categories) that pre-exist and shape action.

Agency — the capacity of people to interpret their situation, choose between courses of action, and act otherwise.

Social fact — Durkheim's term for a way of acting or thinking that is external to individuals, general in a population, and exercises constraint.

Structuration — Giddens' theory that structure is both the medium and the outcome of the practices it recursively organises ("the duality of structure").

Habitus — Bourdieu's term for internalised dispositions that make certain choices feel natural and others unthinkable.

Morphogenesis — Archer's model in which pre-existing structures condition action, and action then reproduces or transforms them over time.

Figuration — Elias' term for a shifting web of interdependent people, offered as an alternative to the "individual vs society" split.

Interpellation — Althusser's term for the way ideology "hails" individuals into ready-made social positions.

Emergence — the appearance of collective-level properties that no individual possesses or intended.

Capability — Sen's term for what a person is actually able to do and be, as opposed to what they are formally allowed.

One — assemble a choice. Take one significant decision of your own and do to it what this lesson did to Meera's. List every fact that was already in place before you had a preference: money, geography, family expectations, what you had seen people like you do, what information reached you, what was available that year. Then ask what was genuinely yours. Most people find both lists are longer than expected.

Two — the cousin question. Find someone from a similar position to yours who made a very different choice. What did they have, know, risk or refuse that you did not? This is the sharpest available test of where structure stops.

Three — measure the room. For one constraint in your life right now, write down what would have to change for it to loosen: a rule, a price, a person's opinion, a law, a skill. This converts a feeling of stuckness into a list — which is what sociological thinking is for.

Four — spot the dial. In the next public argument you hear, identify each speaker's implicit position on structure versus agency. You will find that people move the dial according to whether they are explaining their own success or someone else's failure.