PowerA Radical View

Power is slippery. You can see a vote tally, you can watch a debate, but you can also feel that the real action happened somewhere else—before the roll call, before the cameras. Steven Lukes walked straight into that tension. In "Power: A Radical View," he doesn’t just add footnotes to an old argument; he widens the lens. Start with decisions you can watch. Then look at how the agenda got set. Finally, ask how people’s very preferences get shaped. That’s the move: one dimension, two dimensions, three. Each step expands what we count as evidence, and each step makes power both more interesting and harder to pin down. The first stop is the one-dimensional, pluralist view, the home turf of Robert Dahl. Here, power shows up in concrete fights and visible outcomes. Dahl’s shorthand is memorable: A has power over B when A gets B to do something B wouldn’t have done. Simple, direct, and very measurable. Nelson Polsby later distilled Dahl’s method into something like a scorecard: identify who proposed the alternatives that passed, who got vetoed, and rank actors by how often they win. The focus is on observable behavior—motions made, amendments offered, vetoes cast—and on decisions that leave a paper trail. If you can watch it, count it, and link it to an outcome, it belongs in the analysis. That clarity is a strength. It anchors power in public events instead of hunches about backroom influence. But there’s a catch. If you only study fights that break the surface, you’ll miss the work that kept other fights from surfacing at all. That’s where Peter Bachrach and Morton Baratz crack the story open with what they called the "two faces of power." They argue that power operates not just in decisions but in nondecisions—moments when an issue never makes it to the agenda. Imagine a topic so radioactive for the status quo that it gets defined as "not a real problem" and quietly shelved. That absence is not an accident. It’s power doing preemptive work. Bachrach and Baratz talk about the mobilization of bias: the way dominant values, rules, and routines tilt the playing field so some grievances never ripen into public claims. Here, coercion isn’t only about force. It includes influence—changing someone’s behavior without an explicit say-so; authority—eliciting compliance because the demand looks legitimate; and manipulation—securing agreement without full recognition of what’s at stake. The point is not to multiply labels. It’s to show that agenda control, rule design, and the policing of what counts as "political" are themselves forms of domination. The idea of nondecision needs a little unpacking. A decision is a choice among presented alternatives; a nondecision is the organized avoidance of those choices. It can block a challenge before it emerges or defang it after it does. Bachrach and Baratz emphasize finding which issues are shielded from consideration—declared "settled," "technical," or "out of scope"—even when there’s real risk of conflict. It’s a practical way to look upstream from votes and ask why the river is so calm. And it reminds us that treating quiescence as consent is dangerous. People might not be clamoring for change because the structure has made some demands unsayable or uninterpretable as politics. Lukes takes that critique and pushes it one level deeper. His three-dimensional view says: don’t stop at who decides or even who sets the agenda. Ask who shapes what people come to see as possible, desirable, or natural. If people never express a complaint because they can’t imagine a legitimate alternative, that’s not peace; that’s power working through preference formation. Lukes makes two conceptual moves to ground this. First, he distinguishes overt conflict from latent conflict. Overt conflict is the stuff of hearings and protests—claims clashing in public. Latent conflict exists when people’s genuine interests diverge from the status quo, but institutions, norms, or ideologies keep that tension quiet. Second, he separates subjective preferences—what people say they want—from "real" interests—what affects their well-being. That term "real" is loaded and contested, and Lukes knows it. But it does important work here. It lets us see that the absence of complaint does not settle the question of whose interests are served. Think of this as three axes that intersect. One axis is decision-making and the agenda: who can push or block outcomes, and who decides what gets a hearing. A second is the definition of problems: which potential grievances are recognized as political, and which are downgraded to "private," "cultural," or "natural" matters. The third is the visibility of conflict: open and noisy, masked and coded, or hidden altogether. Once you see those axes, quiescence looks different. Silence might be harmony. Or it might be what happens when socialization, ritual, and institutional routines train people to want what the order already provides. That last point is crucial for Lukes. He’s not saying that persuasion or social learning are spooky forces. He’s saying that power includes the processes by which people internalize norms and interpret the world so that certain demands never feel legitimate. Consent can be manufactured without lies; it can emerge from the way schools, workplaces, media, and procedures make some goals feel reasonable and others feel outlandish. In that world, making someone do what they otherwise wouldn’t is only one kind of power. Another is making them want what you want. When that happens, domination can wear the mask of ordinary order. Now, each layer still matters. The one-dimensional approach earns its keep because politics does, in fact, produce observable winners and losers. If you care about who secured which policy, counting wins is indispensable. Bachrach and Baratz add the agenda, the nondecision, and the mobilization of bias—vital for explaining why some fights never start. Lukes subsumes both and adds preference formation and latent conflict, which helps explain why consent can align with domination. Put together, the three-dimensional view asks three questions in sequence: who chooses among alternatives, who decides what the alternatives are, and who shapes what people think is worth wanting. That compound lens explains why domination can persist without open rebellion and why apparent agreement can mask deep misalignment of interests. There’s also a methodological spine running through this. Dahl and Polsby championed observable behavior: propose, counter, veto, pass. That fits a research design where you pick issues with clear conflicts, track participation, and rank outcomes. It’s clean, but it leans heavily on case selection and visible contests, so it can misread quiet as contentment. Bachrach and Baratz force a redesign. If nondecisions matter, you have to study how agendas get set—the rules for what gets docketed, the procedures that gatekeep proposals, the values that carve topics into "political" and "not." They admit this is empirically hard and, at times, may feel philosophical. You won’t always have a roll call to point to. But hard isn’t the same as impossible. You look for the telltale signs: issues that routinely stall before committees are formed, frames that define structural harms as individual failures, rituals that convert political claims into technical "fixes." Lukes then raises the verification stakes. If you talk about "real" interests, how do you know what they are? He flags the danger directly: expressed preferences can diverge from welfare-relevant interests, and inferring the latter is fraught. That’s not a reason to give up; it’s a warning label. Empirically, he urges researchers to trace the channels of socialization and ideology—how institutions, norms, and narratives shape what people find sayable and thinkable. The presence or absence of explicit conflict doesn’t decide the power question. What matters is whether the alignment of preferences with outcomes emerges from fair consideration among recognized alternatives, or from a structure that made some alternatives invisible. The job is to detect misalignments between what’s publicly articulated and what’s at stake, without pretending we can read minds. That calls for evidentiary humility and layered inference, not a retreat to "only count what you can see." The payoff of the three-dimensional view is not melancholy theory; it’s a sharper map. It keeps the concreteness of visible decisions. It adds the architecture of agenda-setting and the quiet labor of nondecisions. And it brings into focus the background processes—education, media, organizational routines—that prime preferences and produce quiescence. Normatively, it also clarifies the stakes. Autonomy and consent are not just about saying "yes" at the end of a process; they’re about having real alternatives recognized, real reasons weighed, and real interests engaged. Domination can be most effective when it looks like common sense. If you’re designing research under this lens, Lukes’s message is practical. Study the decision moments, but also the political timetable—who can delay until momentum dies, who can demand "more study" until urgency evaporates. Track which problems are named and which stay unnamed. And be candid about the limits: some forms of power will remain unobserved in any single study. That’s not failure; it’s the terrain. Where does this leave us? With a progression rather than a replacement. Dahl is still in the toolkit. Bachrach and Baratz add upstream diagnostics. Lukes wraps those and asks about preference formation and latent conflict. Together, they explain why power is hard to see without widening the circle of evidence. And they give us language for something people intuit every day: that what’s on the table, and what we’re taught to want from it, is already part of the game. In closing, a brief look forward. The three-dimensional view doesn’t demand clairvoyance; it asks for triangulation. Combine behavioral traces—who won what—with institutional analysis—who controlled the docket—and interpretive work—how norms and frames made some demands unthinkable. Keep the claims proportionate to the evidence, and resist the temptation to equate silence with consent or visibility with importance. If power can hide in plain sight, our methods have to look past the obvious without letting imagination outrun proof. That’s the balance Lukes pushes us to strike: expansive in what we count as power, careful in how we say we’ve found it.