Emergent Urbanism

Rediscovering Urban Complexity

Modeling the processes of urban emergence


The growth process of an emergent city actually consists of five growth processes. These processes are hierarchically related, that is to say the morphology decided by processes at higher levels of complexity depends on decisions taken at lower levels of complexity. They are not constrained by one another, as modern planners claim when they clear slums in order to build their architectural vision, but expand upon one another, creating a landscape that is tied to a history of adaptation and transformation in order to meet the needs of the present at every point in time.

Each transformation is the decision of an individual, acting within the context he perceives and the ends that are identified. These ends may be within his own sphere of life, by expanding his home, or subdividing his property to build a home for his grown children, but more likely they are the consequences of identifying a potential created by the individual actions of others. For example, if a sufficient number of neighbors have settled, opening a bakery. In this way networks are built upon the potentials created by the last network extension, (in one such instance by capturing residual movement in the grid as Bill Hillier describes) and the city increases in complexity.

The foundation process of a city, before anyone can even imagine a city being there, I call the place. A place is nothing more than a free surface available to be settled. Newcomers build their home wherever they want in the place, and that implies that they will locate their homes to take maximum advantage of natural features, and space themselves away from their neighbors in order to avoid conflicts over the use of common lands. A place settlement process is how shantytowns are created, except that because there does not exist any functional land ownership in a shantytown there is no limit to how many buildings can be created. Thus the shantytown never reaches the second process of urban emergence, creating a crisis. A place may be created deliberately, by transforming a farm or other types of land use to that purpose, by building fortifications within which land is protected from harm, or a place may be given by nature simply by being available and strategically located.


A place is an open space where people may settle and build randomly

As places become increasingly dense, the use of space by neighbors will create conflicts of proximity. Land will no longer be superabundant. In order to resolve these conflicts a process of land enclosure delimits the boundaries between neighbors' households by negotiating the boundaries of land that is in private and common use. Streets and blocks thus appear, and those spaces where common use is particularly intensive, because of highly valuable natural features or central locations, become recognized as public squares and greens.


Enclosures delimit private and public spaces, and the pattern of streets, blocks and squares emerges.

With available land to settle either enclosed or occupied by public activities, it becomes more difficult for new growth to take place. New buildings built on remaining place must be justified before a community increasingly protective of the remaining open space. In most cases it is much simpler to ask one of the members of the community to give up a part of his property in order to grow the new part of the town, introducing the process of subdivision. These subdivisions are negotiated case-by-case and thus adopt random sizes and shapes, creating a fractal distribution of lot sizes over a long timeline. Some subdivisions split the land into shared courtyards and cul-de-sacs that are administered under a co-property agreement (they never need to involve the community as a whole).


Properties are subdivided to make room for new growth and new network relationships now that open land is in short supply.

Eventually crowding becomes problematic at the same time as the scale of network growth is increasing due to higher population densities. This creates the opportunity not only to open new places to settlement, but also to connect the central city to these new places by a place functioning at a greater scale, near a road or highway, and that provides an encircling bypass around smaller-scale neighborhoods. This is the grid process. This new construction opens up land to construct large market and industrial businesses that are simultaneously a buffer between smaller-scale places and roads but also an integrator of these places into larger-scale networks.


The grid integrates mature places into a larger network of places, and creates new spontaneous development opportunities.

The last process takes place when a large city with many places integrated by many scales of grids develops a mass transit system that becomes more reliable than private transportation systems. When that occurs the need for private transportation falls radically and it becomes possible to live at the centers of this mass transit system without any private transportation, thus radically reducing demand for space. Parking lots can be built over and turned into undifferentiated buildings providing standardized living spaces that can find their match in the very large population. This radically higher population in turn creates a very wide potential for new differentiated networks, and the construction of large buildings is accompanied by many new, differentiated small buildings. This is what enables a place to achieve high density complexity, and we can call it the metropolitan process.


A small number of larger new buildings accompanies a large number of small new buildings resulting from the reduction in space needed for transportation.

A model such as this one is not meant to be a design to be implemented in reality. It serves only as an illustration of the processes, the means through which decisions are achieved, that generate the structure of cities. If we want to do the morphology of an existing city, it is these processes that will help us explain what decisions led to the city's present form. These processes also help us predict the future of the model of urban development we choose to adopt. As an example I have become highly critical of measures that seek to increase the density of subdivision developments by smart growth zoning regulations. They tend to leave the structure of neighborhoods in such a state that further subdivision processes within its tissue are impossible, and the neighborhood becomes unable to adapt itself as its population changes. Instead we should be building low density subdivision developments that can grow naturally into metropolitan neighborhoods, and this growth will be controlled by its community as its members make the decision to give up a part of their property to accomodate the changes the community is undergoing.


I agree until you get to the final hit on smart growth. Modern zoning regulation prohibits the natural sort of subdividing you see as beneficial. Only a transect based zoning system, or no zoning at all, has the robustness to reform itself over time.

An interesting model, but I hope you have further explanation for your prescription at the end.

"[a] Instead we should be building low density subdivision developments [b] that can grow naturally into metropolitan neighborhoods . . ."

The current car-oriented suburbs satisfy [a] but most probably do not satisfy [b].

I highly doubt that such a community would control growth in such a way as to lead to the dense urban development that we find in the older core of European cities that people find so attractive.

What restrictions should the community be making to retain public spaces and prepare for future development?

It is incorrect to believe that the car oriented subdivision is low-density. In most cases there is not enough space left over for any further subdivision of plots. In the latest, smart-growth driven subdivision projects you are essentially getting a townhouse with identical neighbors, only there is no town anywhere to be seen around it to mitigate this downside. It is density with none of the benefits associated with urban density. Since things are already at their worst, further increases in density won't be received positively.

The condition for an emergent urbanism project is whether the process can repeat itself within the forms that are set in the previous iteration of development. All of the processes in the model above can loop and feed back on themselves, and thus they are all complex and fractal-generating. They have, however, scale limits. If those scale limits are not respected in development, the feedback won't take place and there won't be any emergent urban complexity.

I understand that people like high-density cities. I like them as well. But forcing high-density cities on a process that is not meant to create them, the subdivision process, only serves to make the process worse and not better. It makes enemies of land developers who see the absurdity of the effort, and creates another showcase of the failure of repression to force great cities to exist, and worst of all, spreads horrible type 2 linear process geometry on the landscape.


If I understand your points correctly, I agree with you.

But if I am a developer or a city planner with a parcel of property ripe for development, what conditions or restrictions should be made so that a desirable dense urban development can emerge over time? That is a question that interests me.

I wonder whether you have discovered some basic rules that would not unduly burden today's developer, but would provide conditions that would allow for future fractal-generating development that eventually (100 years, 200 years?) would yield the high-density development that is so appealing. Developers in the U.S. often already work within numerous restrictions on lot size, frontage, etc. which as you noted do not create the necessary conditions. Even when the development is characterized as "smart growth."

Rather than force developers to build high-density mixed-use developments, ask them to build the low-density developments that they seem to want to build, but with conditions that allow and encourage future development that leads to a higher density organically.

Maybe that means setting aside a portion of the parcel for a community center that can become a retail center and eventually a town square, for example. I don't know if or how that would work, but that's one possibility.

What do you think?

Further comment

Please send your comments by email at mthl@mthl.info, or find me on Twitter @mathieuhelie. The commenting system is closed at the moment as no measures can hold back blog spamming bots.