The Three Faces of Power and Beyond

Power has run under this entire course — in class (Part 8), in institutions (Part 9), in control (Part 10) — and it has been deferred until now, because it is the hardest concept in the discipline to define, and the reason it is hard is itself the most important thing about it.

The trouble is this: the more effective power is, the less it looks like power. The power you can see — the order given, the threat made, the conflict won — is the crudest and least interesting kind. The power that matters most is the power that produces outcomes without any visible exercise at all — that keeps things off the agenda, that shapes what people want, that makes an arrangement seem so natural that no one thinks to contest it. This lesson is about learning to see the power that is designed not to be seen.

A town with a poisoned river and no protest.

Imagine a town built around a single factory, on a river the factory pollutes. The pollution is real; the townspeople's health suffers; by any measure they have a grievance.

And there is no protest. No campaign. No demand. The issue never comes up.

Now ask: what is the state of power in this town? There are, in fact, three completely different situations that would look identical from the outside — three faces of power, and telling them apart is the whole of the lesson.

First face. The townspeople raise the pollution at the town council. They lose the vote. The factory owner's allies on the council outvote them, or the owner threatens to close the plant and cost them their jobs, and they back down. Here power is visible: there was a conflict, a decision, and one side prevailed over the other's expressed preference. This is power as ordinarily understood — A gets B to do something B would not otherwise do — and it is the only face most accounts of power can see.

Second face. The townspeople want to raise the pollution, but it never reaches the council at all. The agenda is controlled — the owner's allies chair the committees, the local paper will not run it, the procedure for adding an item is impossible to navigate, and everyone knows that raising it is futile and possibly costly. So the grievance exists and is never voiced. Here power operates by keeping the issue off the agenda — by ensuring the decision is never made, because a decision that is never made cannot be lost visibly. This is a second face of power, and it is invisible to anyone who studies only the decisions that get made.

Third face. The townspeople do not raise the pollution because it does not occur to them that they could. The factory has always been there; the river has always been like this; the town's prosperity depends on the plant, and to be against the plant would be to be against the town, against your neighbours' jobs, against yourself. They have come to accept the pollution as natural, inevitable, the price of a living — and they do not experience it as a grievance at all. There is no conflict, not because power was exercised in any visible moment, but because power has shaped what the townspeople want and believe — their very perception of their own interests.

Three towns. From the outside, all quiet. In the first, power won a fight. In the second, power prevented the fight. In the third, power ensured the fight was unthinkable — and the third is both the most effective and the hardest to prove, because there is nothing to point to. No decision, no suppressed grievance, no smoking gun — only a quiet town by a poisoned river, and people who do not want what, on some accounts, they should.

How do you study a power whose success is that nothing happens?

This is the methodological heart of the concept and the hardest problem in it.

The first face is easy to study: you look at decisions, identify the conflicting preferences, and see who won. This is why so much political science studied exactly this — it is observable, measurable, and it fits the evidence discipline of Part 7.

The second face is harder but still tractable: you look for grievances that exist but do not reach the agenda, for the mechanisms that keep them off, and for what a group would demand if the barriers were removed. It requires studying non-decisions — the things that do not happen — which is difficult but not impossible (see 7.3.2 on studying what is absent).

The third face is the deepest problem in the discipline. To claim it, you have to say that people's actual preferences differ from their real interests — that they want something other than what would truly benefit them, because power has shaped their wants. And that is a dangerous claim (see 8.1.3 on adaptive preferences, 5.5.1 on hegemony, 8.2.3 on individualisation — the same problem, arriving from four directions).

Dangerous because it seems to license the analyst to overrule people about their own interests — to say "you only think you want this; your real interest is what I say it is." That is condescending, potentially totalitarian, and it is exactly the objection the concept has to answer. How can you ever establish someone's "real interest" against their expressed preference without simply imposing your own values (see 7.1.2)? The lesson has to hold the insight — that power shapes wants — without the arrogance that the strong version invites. That balance is the whole difficulty, and it is not fully soluble.

The three faces, precisely

The classic sequence, and what each added.

The first face — the pluralist view. Power is the capacity of A to get B to do something B would not otherwise do, revealed in decisions where preferences conflict. Studied by looking at who prevails in observable conflicts over key issues. Its virtue is rigour — it studies what can be observed. Its limitation is that it sees only the fights that happen, and misses everything that never becomes a fight.

The second face — agenda control. Power also operates by keeping issues off the agenda — by the mobilisation of bias, the set of arrangements (procedures, rules, values, institutions) that systematically benefit some and disadvantage others by determining what gets decided at all. Non-decision-making — the suppression of challenges before they reach the point of decision — is an exercise of power, and studying only decisions misses it. A grievance that never becomes an issue is a victory for whoever benefits from the status quo, and it is a victory won without a battle.

The third face — the shaping of preferences. Power operates most effectively by shaping people's very wants, perceptions and beliefs — by securing their consent to an arrangement that harms them, so that no grievance forms. This is the most insidious face because it prevents conflict from arising, and it works best when it is least visible. "Is it not the supreme exercise of power to get another to have the desires you want them to have — to secure their compliance by controlling their thoughts and desires?"

The three are cumulative, not competing. A full account of power in the town needs all three — the votes that were lost, the issues that never reached a vote, and the grievances that never formed. And they are ranked by effectiveness in inverse proportion to their visibility: the first is the weakest and most visible, the third the strongest and least visible. Which is the counterintuitive core of the whole concept — the best power leaves no trace.

And the fourth face, which reframes the other three.

Foucault's account of power (see 5.7.2, 10.3.1) is sometimes called a fourth face, and it is different in kind.

The three faces above all assume power is something someone has and exercises over someone else — A over B, the factory owner over the town. Foucault's power is not possessed but exercised through relations, dispersed through the whole society, productive rather than only repressive, and operating through the constitution of subjects themselves.

On this account the town is not a case of a powerful owner suppressing a powerless populace. It is a case of a whole configuration — the discourse of jobs and prosperity, the identities of "worker" and "provider", the knowledge of what a river is for, the norms of what a good neighbour supports — that produces townspeople who want what they want. There is no single holder of power pulling strings; there is a field of power relations that constitutes everyone in it, including the owner.

Its insight: it captures how the third face actually works — not as manipulation by a villain but as the productive operation of discourse and knowledge (see 5.7.2). Its problem: if power is everywhere and constitutes everyone, it becomes hard to say who is dominating whom, what resistance is, or how anything could be otherwise (see 10.3.1 — the same criticism). A theory in which power is a field with no holders struggles to distinguish the owner from the townspeople in the ways that matter to their lungs.

The mature position holds both: the three-faces account is better for asking who benefits and who is harmed (it keeps the owner and the town distinct); the Foucauldian account is better for asking how consent is produced (it explains the third face without needing a puppet-master). They answer different questions, and a full analysis uses both.

The false-consciousness trap, and how to avoid it.

The third face requires distinguishing real interests from expressed preferences, and the danger is imposing the analyst's values as the "real" interest. This is the objection that has to be answered, and it cannot be waved away.

Three disciplines keep the concept honest.

One — the counterfactual test, made as concrete as possible. The claim that power shaped preferences is a claim that people would want differently under different conditions — if they had the information, if they were free from the constraint, if the alternative were available and thinkable. So look for what people want when the conditions change — when a factory in a comparable town does clean up and the townspeople there are glad, when information becomes available and preferences shift, when a movement makes the unthinkable thinkable and grievances that were latent become manifest. The third face is testable in principle: it predicts that removing the constraint changes the want (see 8.1.3 on the Kerala health example — the same test).

Two — humility about the specific case. It is far easier to establish the mechanism (that power can shape preferences — which is demonstrable) than to establish that it did so in a particular case against a particular person's expressed view. Claim the mechanism confidently; claim the specific instance tentatively (see 7.9.1).

Three — the people's own reflection. The strongest evidence for the third face is not the analyst's judgement but the moment when people themselves come to see their former acceptance as shaped — when the townspeople, later, say "we didn't realise we didn't have to live like that." The consciousness-raising, the movement, the changed frame (see 11.2.2) that lets people rename their situation — that is the third face becoming visible from the inside, and it is better evidence than any outsider's diagnosis.

What this rules out: the lazy use of "false consciousness" to dismiss anyone who disagrees with the analyst (see 8.6.2 on the same condescension). What it preserves: the genuine and demonstrable insight that power shapes wants, held with the humility the danger requires.

Because the power you can see is the least of it, and a citizen trained to look only at visible conflict will miss the arrangements that matter most.

Three questions for any situation of quiet.

Is this consent or is it the absence of an alternative? When people accept an arrangement, ask whether they endorse it, whether they have no way to contest it, or whether they cannot imagine contesting it. These are the three faces, and they require completely different responses — the first is legitimate, the second is a blocked agenda, the third is a shaped consciousness.

What is not on the agenda, and who benefits from its absence? The most important political question is often not "who won the debate?" but "what never became a debate?" (see 9.6.1 on agenda-setting — the media version of the same power). Follow the non-decisions — the issues that never arise, the options never considered, the alternatives that are unthinkable — because that is where the second and third faces live.

And would people want differently under different conditions? For any preference that serves the powerful, ask what would happen to it if the information, the constraint or the horizon changed. If it would dissolve, it was shaped; if it would persist, it was genuine — and the difference is testable, in principle, by comparison and by what happens when movements change the frame.

One closing observation that sets up the whole Part. The quiet town is the normal condition of most power most of the time. Domination that has to be enforced by visible power is domination in trouble — expensive, contested, and unstable (see 10.4.1 — order held by force is order that is failing). Domination that has secured consent, that has kept its harms off the agenda, that has shaped its subjects to want what it needs them to want — that is domination that is working, and it produces no protest, no conflict, and no headlines.

Which means the task of a sociology of power is to make the quiet audible — to recover the grievances that never formed, the issues that never arose, the alternatives that were made unthinkable. The rest of this Part is about what happens when the quiet breaks: when grievances form, agendas are forced open, the unthinkable becomes thinkable, and people who were supposed to accept their situation decide, instead, to change it. That breaking is the subject of social movements (11.2), and it begins with the hardest question of all — not why the quiet breaks, but why, given how much there is to protest, it breaks so rarely.

The more effective power is, the less it looks like power — which is the counterintuitive core of the concept. A quiet town by a poisoned river has three possible states, identical from outside: power won a visible fight (first face), power kept the issue off the agenda so no fight occurred (second face), or power shaped what the townspeople want so the grievance never formed (third face). Ranked by effectiveness in inverse proportion to visibility — the best power leaves no trace.

The first face — pluralist — is power in observable decisions where preferences conflict; rigorous but blind to the fights that never happen. The second face — agenda control, the mobilisation of bias, non-decision-making — keeps issues off the agenda, a victory won without a battle. The third face — shaping preferences, perceptions and beliefs — secures consent so no grievance arises, the most insidious because it prevents conflict.

The fourth face — Foucault's — is different in kind: power not possessed but exercised through dispersed relations, productive, constituting subjects themselves. It explains how the third face works without a puppet-master, but struggles to say who dominates whom — so the three-faces account is better for who benefits, the Foucauldian for how consent is produced.

The third face requires distinguishing real interests from expressed preferences — the false-consciousness trap — which risks imposing the analyst's values. Three disciplines keep it honest: the counterfactual test (power that shaped a want predicts the want changes when the constraint is removed — testable, as in the Kerala example); humility about the specific case (claim the mechanism confidently, the instance tentatively); and the people's own reflection (consciousness-raising, when people come to see their former acceptance as shaped, is the best evidence). This preserves the demonstrable insight that power shapes wants while ruling out the lazy dismissal of anyone who disagrees.

And the quiet town is the normal condition of power: domination enforced by visible power is domination in trouble; domination that has secured consent produces no protest. The task of a sociology of power is to make the quiet audible — and the rest of the Part is about when the quiet breaks.

The first face of power — A getting B to act against B's preference, observed in decisions over conflicting interests.

The second face of power — agenda control; keeping issues from being decided at all.

Mobilisation of bias — the arrangements that systematically determine what gets onto the agenda.

Non-decision-making — the suppression of challenges before they reach a decision; an exercise of power.

The third face of power — the shaping of wants, perceptions and beliefs so that no grievance forms.

Real interests versus expressed preferences — the distinction the third face requires, and the source of its danger.

False consciousness — accepting an arrangement against one's real interest because power has shaped one's wants; genuine as a mechanism, dangerous as a verdict on a particular person.

The fourth face (Foucauldian power) — power as dispersed, relational, productive, and constituting subjects; explains the third face's mechanism.

The counterfactual test — the claim that shaped preferences change when the constraint is removed; the discipline that makes the third face testable.

Latent versus manifest grievance — a grievance that exists unvoiced, and one that has become an expressed demand.

One — find the three towns. Take one arrangement people accept quietly and ask which of the three faces holds it — endorsement, blocked agenda, or shaped want. The three require different responses.

Two — look for the non-decision. For one political situation, ask not who won the debate but what never became a debate. That absence is where the deeper power is.

Three — apply the counterfactual test. Take a preference that serves the powerful and ask what would happen to it if the information or the constraint changed. If it would dissolve, it was shaped.

Four — catch the trap. Find a use of "false consciousness" or "they don't know their own interests" and ask whether it passes the counterfactual test or just imposes the speaker's values. Most fail.

Five — recover a formed grievance. Identify something now widely felt as an injustice that a generation ago was accepted as natural. The change from latent to manifest is the third face becoming visible.