Can architecture be democratic? Most people would readily agree that the built environment is bound to be political. Yet in the popular imagination the combination of “architecture” and “politics” tends to conjure up distinctly undemocratic figures: totalitarian leaders designing monumental edifices and avenues for eternity. And if authoritarians fancy themselves as architects, so a certain logic goes, architects often act like authoritarians: at best they might create something for the people, but not anything meaningfully seen as of the people and certainly not by the people. And yet there are ways of judging architecture and space to be more or less democratic — and, to some extent, practical strategies to render architecture (and also urban planning) democratic. These are bound to become ever more important in the twenty-first century, as our age is one of unprecedented urbanization and hence new planning and building challenges around the globe.

At first sight, the obvious way in which architecture and urban planning should be democratic is by involving citizens in the planning process. This is easily said, but exceedingly hard to do. The reasons are not just practical ones, such as: how do we get everybody into the same room at the same time? For who exactly should be included in forms of participatory democracy in the first place is far from clear. A typical candidate for a criterion of who to include is the notion that all those affected by a building decision should have standing to enter the planning process. But who is affected? And who decides who counts as affected? The inhabitants of a building (or “users”), to be sure — but what about neighborhoods, a city as a whole, or an entire country? Might it even be, for symbolically highly charged sites such as Ground Zero, nothing less than the world population? 9/11 undoubtedly was a global event, but does that make every human being part of the constituency for decisions about sixteen acres in Lower Manhattan?

Political theory has no straightforward answer to such questions; there are no agreed criteria for deciding who should count as “affected.” Think again of Ground Zero, for many years undoubtedly the most closely watched piece of land in the world. There were carefully planned exercises in participation, branded as “Listening to the City,” to involve ordinary New Yorkers through a series of moderated conversations. But the relatives of the victims of the terror attacks, not without good reasons, argued that their voices should count for more in the design decisions. Even when it comes to elected officials, the questions of authority and standing are not easily settled. For instance, should Michael Bloomberg, as mayor of New York City, have had more of a say than the governor of New York State, George Pataki, whose voice was in the end decisive for picking the master planner of the site, the architect Daniel Libeskind?

Libeskind himself — who at one point called the process leading up to the construction of the new World Trade Center complex “the most democratic in the world” — added another possible understanding of the relationship between architecture and democracy: namely, that it’s not so much the process as the architectural and spatial “product” which ultimately has to be democratic. Libeskind — a self-declared “people’s architect” — always insisted that he wanted to create “a space for people, not just corporations.” But what is a “space for people”? One possible answer might be: a space where citizens recognize their polity (and themselves) as subscribing to democratic values. Particular shapes or even materials are often said to symbolize such values: glass, for instance, is taken to signify “transparency” (and, by implication, democratic “accountability”). Greek and Roman statues are supposed to remind citizens of democratic ideals from the ancient past. And Libeskind himself chose the symbolic height of 1776 feet for what used to be called the “Freedom Tower” (again, Pataki’s choice, although never made official).

The problem is that such representations of democratic values are not equally comprehensible or, if you like, accessible. To the innocent observer, it needs to be explained that the tower is 1776 feet high and perhaps also for what that number is supposed to stand (and perhaps also that the spire had once meant to be on the edge of the tower’s roof and echo the arm and the torch of the Statue of Liberty). Many well-intentioned attempts to make buildings symbolic will simply not be understood at all. Who remembers that Minoru Yamasaki, the architect of the original World Trade Center, thought the twin towers represented “man’s belief in humanity” and “his need for individual dignity”? Such declarations might demonstrate the hubris of the architect, but do little to influence the way a building might or might not become part of a lived political experience.

More important, a symbolic gesture towards democracy can literally go together with a denial of access: it is perfectly nice that German citizens can watch their deputies from Norman Foster’s glass dome on the re-designed parliament building in Berlin — they, as the ultimate sovereign, are symbolically elevated above their representatives. (The parliament building in Canberra follows a similar logic.) But de facto they remain distant spectators — they cannot even hear what is being said in their name. Representing democracy and facilitating democracy are two different things; and what one may well dismiss as “democratic decoration” (“Look! Glass! It’s democracy!”) can actively distract from noticing the actual absence of democratic facilities.

That leaves two perhaps less obvious — but actually quite coherent — ways to understand the relationship between architecture and democracy. One amounts to an ethical and political imperative that architects should not design for authoritarian regimes. Of course, architects will always be tempted to say that they are not responsible for the context of their work. Or they might even claim that their buildings have a politically subversive effect, or that the meaning of even the seemingly most totalitarian structures can be changed and re-coded by the people themselves. Think of some Nazi architecture in today’s Berlin: it houses the offices of a liberal democratic government — and at the same time reminds citizens of the need to reflect on the totalitarian past. In fact, one might even say that if democracy is ultimately about an ideal of collective autonomy, then such collective re-signification over time is the real thing, or, at least, a process that’s much more important than adding a bit of glass here and an allusion to heroic Founding Fathers there.

Yet such a relaxed attitude towards the political status of architecture’s actual, paying clients overlooks that “starchitects” who today effectively sell their name (or brand recognition) to authoritarian regimes in Central Asia — think of prominent, even iconic, buildings by Zaha Hadid and Norman Foster in Kazakhstan and Azerbaijan — directly serve the strategic goals of these regimes. Petro-dictators like Ilham Aliyev crave cultural prestige and international recognition; “starchitects” deliver both and hence shore up the legitimacy of governments that allow them a freedom to build (and raze and displace) that can rarely be had in democracies. Libeskind, who has famously refused to work for despots with their “homogeneous systems” that make it easier to realize one’s plans without messy political compromises, rightly insists that one cannot separate “formal geometry from the context of who they were commissioned by and the morality of those states.”

But which “morality” do democracies offer, by contrast? One, I would say, that is precisely not set in stone; democracies do not pre-determine political meanings, but the structures they offer for the creation of meaning are also not random. Democratic architecture provides accessible space (whether public or private) where citizens can freely gather, debate, and — above all — protest. What recent political events, whether the Arab Spring or social upheaval in Southern Europe, have made very clear is that even in the age of the Internet and social media, actual physical space still matters a great deal. Democratic architecture should make room for what the political scientist John Parkinson has called “democratic performance,” enabling citizens to create their own political messages.

What does this mean concretely? Parliaments should have large, empty spaces on one side, where the people can gather to voice concerns (what is sometimes called “Authorized Assembly Space” — even if that term has slightly Orwellian overtones). These spaces, as Parkinson has underlined, should not be overly landscaped: filling them with trees and benches might make them more pleasant for tourists, but less usable for what Parkinson has called “purposive publics.” To be sure, purposive publics are not the same as “the people” themselves — but the latter can never fully appear in a democracy, nor can they be fully represented without the possibility of contestation. As the late French political theorist Claude Lefort always insisted, the logic of representation in a democracy is very different from that in a monarchy: the king can represent the realm without remainder; but, in a democracy, the place of power remains always empty and contested. What people (never “the people”) can do, though, is, in the American example, “march on Washington.” To make parts of the people visible — especially those excluded or forgotten — is part of the democratic function of public space; hence the avenues need to be clear, and the room needs to be there, to march and assemble in a capital.

Parliaments, then, never fully represent “the people,” but neither do citizens gathered in a public space. The German political scientist Philip Manow has suggested that the reflection of representative buildings in a pool (as in “reflecting pool”) nicely symbolizes the fluid, impermanent nature of democratic representation: neither the assembly nor assembled citizens ever fully represent the people without remainder; they are fleeting and, on occasion, they might seek to disturb or even de-legitimate each other. It’s by making that dynamic, inherent in democracy, visible and, hopefully, productive, that architecture and urban design can both symbolically represent and facilitate democracy.

Of course, one might object that this argument makes democratic architecture into a matter of official sites only, even if these sites can be contested, instead of emphasizing the character of democracy as an everyday practice (no-one marches to Washington every day). The most plausible — but pragmatic, not philosophically watertight — answer one can give here is that, as theorists like Giancarlo de Carlo had argued all along, planning processes in general should indeed be open for public input. The best criterion for selecting the relevant public remains the principle of territoriality; or, put in plain terms: who resides in the neighborhood prima facie has to count as affected and has to have standing. However, with larger policy questions — about gentrification, for instance — democratic publics as a whole will need to get involved (and protest, if necessary). Planning policies create urban patterns — patterns, which, like patterns of wealth distribution, raise fundamental normative questions about collective self-determination; they cannot just be decided locally. They are questions about justice and degrees of diversity, for instance, and they will need to be made subject to the widest possible contestation in a polity. The rest, then, is not a matter of democratic theory, but actual voices of citizens.