Emergent Urbanism

Rediscovering Urban Complexity


Having read and been influenced by A City is not a Tree some years ago and by some of the writing of Nikos Salingaros, I have always wondered whether the same topology and algebra can really be used at the level of the building, of the city, and of social networks. I make the simple observation that within the past century we have gone from a grid pattern for streets and elegant hierarchical arrangement of elements at different scales for buildings to a grid pattern on buildings and elegant looking road hierarchies.

And they are only simple examples, but a lot of fractals, including Sierpinski carpets and the Mandelbrot set are actually hierarchical. Show me a city based on a Julia set with an infinity attractor and I'll say "uncontrolled sprawl", but pretty from the air.

So while a building facade probably gains from being scale invariant to a degree and having a bit of a fractal dimension, I'm not convinced that the same necessarily holds at the urban scale. For one thing, social networks are no longer as localized as they once were, and I would refer you to Barry Wellman's work on locality. Urban space is only locally metric, and measures of distance vary according to means of transportation and communication. Stores and friends can be close by foot and far by car or vice-versa based on several factors controlled by planners. But Facebook and eBay are always the same distance.

A city is always responding to changing conditions, but I am not certain the response necessarily restores an equilibrium. So for instance a random fluctuation can create a local demand for a rendering plant or a bagpipe school, things that have a negative impact on neighbours. Satisfying demand, in the absence of planning, can create its own dislocations, with people who can afford it gradually moving away over a period of years. Emergent phenomena are not necessarily static solutions even long after the perturbation is gone. That's not necessarily bad, but what I don't like is the cyclical theory of urban development, where it is believed that a community must fall into decrepitude and then be redeveloped as a natural cycle. This is only applied to dispossess the poorest communities and hand them over to developers. This type of feedback cycle (poor=bad, rich=good) tells me something is broken in the system, where a drop in land prices through the mechanism of human misery is required in order to restore overall equilibrium.

Thanks for giving a clear view on the fundamentals of this.

Mathieu, have you considered binding your essays up and submitting for pamphlet architecture?

Matthieu, I started reading the blog a couple months ago and always find your articles refreshing and provocative. As a RE banker who finances large developers, I think most of the criticism of modern real estate development industry is spot on. Just a couple of my own observations , keeping in mind I'm not a planning professional-

I think what Patrick is getting at is that you need to recognize how the transition from merchant/agrarian capitalism to industrial capitalism played a big part in altering the process of urban growth. Changes in the scale of production and the concentration of capital did allow for the "natural" emergence of factory towns and other forms of dictated development. They were natural in the sense that they grew in the absence of government planning. Yet your narrative starts more than a century later with the introduction of the first modern suburban subdivision (or so I assume that's what you are referencing). By doing this you pass over a lot of history where government planning evolved in response to correct the pathologies generated by industrial development. This is unfortunate, because I suspect that this period represents an instructive middle ground for planners and developers trying to find the right trade-off between spontaneity and predictability.

The Birmigham development process that Patrick describes sounds like a decent blend of the two, which is to say it allows for much more spontaneity than your typical master-planned subdivision. The developers (and builders subsequently) did not have to zone the land, get planning commission approval, defend their plan at public hearings, apply for permits, conduct environmental impact assessments and traffic studies, and pay hefty "impact fees" to grease the palms of local officials. And as CG points out, the very idea of starting a new town based on a grid, then allowing the land developers to sell to builders who had mostly free reign is a foreign concept to modern town planning regimes.

Any disagreements aside, I really appreciate the level of dialog you promote here, Mathieu. It is quite unusual for an author to respond so thoroughly and respectfully to comments, but it is very valuable. Thanks.

This is an interesting post (the comments as well) but I think you miss some important societal trends that were based mainly on technological innovation. Most importantly, the way people shared information now is much closer to the way it was shared before the advent of homogenized and mass produced TV and radio.

Twitter, facebook, blogs, etc... are a technologically advanced form of a 18th century town hall meeting or conversation over a beer. Contrast that with 20th century TV and radio (and even most of the newspaper) which is much more like listening to a speech, or rather, speech after speech after speech.

Modern suburbs don't lack fractal or emergent qualities, they are just much much simpler than their predecessors because fewer voices were involved in crafting the rules.

In the 50s, people bought into the big yard and white picket fence not because that was the most wonderful thing, but because it was cheap for developers with cheap land outside cities to control public opinion.

As we begin to add new rules--informal or otherwise--and not merely add complexity to the existing ones, we have a chance to alter the entire development process. We can promote pedestrianism or sustainability, health or investment.

There is no strict dichotomy between a spontaneous and planned city. There is a spectrum. As I've argued before, every city needs to have an emergent dimension at some scale in order to continue functioning through environmental change. While the completely random city of the middle ages was fully emergent, from the shape of the landscape to the details of buildings, early city planning efforts mostly limited themselves to define a landscape structure such as an orthogonal street grid.

A very significant threshold was passed, however, when the land developer as a subdivider was transformed into the land developer as real estate salesman. Subdividing land has always been part of the process of urbanization, and many of the patterns of spontaneous cities can be explained by property owner subdividing their lots to increase the density and benefit from the higher demand for urban land. 19th-century regulations obliged land subdividers to uphold street grids, but the idea of "infrastructure" was not well-developed, and in fact streets were typically nothing more than dirt paths that required no investment.

It was the advent of infrastructure and municipal utilities that forced land subdivision to become today's development industry. Municipalities wondered how the extension of utilities and public services were to be paid for. The solution was to impose the costs onto developers, who were required to build infrastructure to regulation and turned to the banks to finance this additional work. What it meant was that the growth of the neighborhood had to be rapid enough to pay for all these utilities and it could not exceed the density for which the development was planned, which would impose higher costs on the municipality once the developer had long moved on. The developer therefore had to sell a lot of homogeneous houses very quickly, and those houses remained there ever since.

This is not to say that large-scale real estate development did not exist prior to the 20th century, as Patrick points out. But the concentration of capital necessary to realize them was quite rare during the 19th century, available at first only in England and then spreading to other countries gradually. The story of the renovation of Paris involves large banking oligarchies partnering with the prefecture to build new streets, and while these streets today are considered to be some of the most beautiful in the city, built with all the craftsmanship of 19th-century industrial production, they are not the most alive. They succeeded because they were inserted within the tissue of a spontaneous city and therefore came to complement it. New neighborhoods in the suburbs did not have this advantage, and unless they were close to a traditional village, still to this day do not have much life of their own.

In North America capital concentration came even later, the Federal Reserve System founded in 1913 and the Bank of Canada in 1935. Without this funding the modern system was not possible, and people who wanted to borrow to build a house had to borrow the savings of their neighbors through the local small-town bank.

Have been reading this blog for a while, and always enjoy your posts. Interesting comments by Patrick as well -- I can't speak much for Birmingham, but in my city of Nashville the process of development, from the 1750s to the 1950s and later, also involved a large degree of top-down planning.

The main differences from today were that 1) neighborhoods were built on tight grids around some existing infrastructure, rather than as isolated greenfield developments, and 2) although large developers bought farmland, subdivided it and sold it as early as the 1850s (even the CBD was created through a similar process in the 1780s), they did not actually build the houses themselves. My own street, very near downtown, was platted in the 1880s, with the first house completed in the 1890s, and the last vacant lot built on only in the 1960s. I think this is all fairly typical of the older portions of American cities.

While it's certainly not anything like a spontaneous process -- zoning laws would have stopped that 80 years ago regardless -- the responsibility for the bulk of the built environment was nonetheless on the individual. You can still see evidence of this today in the dedication plaques on old commercial structures, which contain the name of an owner and the year the structure was built, indicating pride in the building and in individual ownership. The last one I have seen downtown is dated 1957.

Patrick makes an excellent point, but I think the differences in the pre-1950s pattern of urban development are fairly significant. I agree, though, that it would be going too far to imply that the the old process was "spontaneous."

Nice post. I really enjoyed this and look forward to the next installment.

I do take issue with a couple statements, however.

"But for smaller towns the same project can double the size of the urban fabric." This seems unlikely. Do you have evidence of this actually happening?

"The adoption of mass-production processes, or development, in substitution for spontaneous urban growth in the mid-20th century created for the first time a phenomenon of alienation between the inhabitants and their environment. " Again, this seems like pure speculation. It hardly seems likely that no one was alienated from their environment prior to 1950. Consider the factory towns of the 19th century, especially the giant dorms of New England factory towns. These were not constructed incrementally by individuals, but were built en masse by factory owners.

One theme I notice in many of your posts is the idea that the planning/development system shifted from an individual-based system to a corporate-based system, dominated by developers, in the mid 20th century. But you never explain why you think this, and you seem to ignore the fact that developers played a large role in development prior to 1950 as well.

For example, Birmingham, Alabama, my home town, was developed in the 1870s and 1880s by industrial developers around steel mills. The land for the entire city was purchased by developers and subdivided into a system of blocks by the developer. So the developer owner the land. The developer subdivided the land. The developer sold the land. Individuals purchased the land and constructed the buildings, but they likely needed bank loans to do so.

So, while development happened more incrementally, banks and developers still controlled a great deal of development. This doesn't look much like the spontaneous slums you write so fondly of. And it seems unlikely that any significant permanent development in the US and Western Europe has happened spontaneously since the middle ages. Can you address this?

If you don't want them living in a floodplain or in polluted waste, aren't you pretty much arguing in favor of city planning? Are you not implicitly agreeing that there needs to be some sort of system to ensure that people aren't living in floodplains or in polluted waste meaning that system of land allocation needs a fair amount of regulation?

Is the problem in Sacramento one of land use regulation or the inadequate provision of social services?

If Sacramento was to deregulate land use planning the consequence would be that people are living on polluted land or living in a floodplain in substandard housing.

There is alternative solution, maintain the current planning regime, but increase social services to the populations most likely to be homeless.

Essentially that is what Sacramento County is doing. Arguably the plan should be implimented faster. But I think the outcome will be a lot better than what you are advocating.


You are still missing the point Ed. I don't want these people to live in flood lands or polluted waste. But if you kick them off that land, where are they going to live? There's a reason they ended up there in the first place. The city has a responsibility to regularize them somewhere, to allow them some place they can live their life. The city has not done so because it only does business in collusion with developers.

The reason people live in squatter camps is that they were forced to the margins, and the city has a responsibility for taking them back into normality. That means that it has to abandon its planning system of city and developer partnership, because squatters are their own developers and can't afford the expense of the planning system.

The tent city that appeared on the Oprah Winfrey show as in Sacramento. That tent city appeared on land owned by 4 entities, the Blue Diamond Almond Growers Association, Southern Pacific Railroad, Sacramento Municipal Utilities District and the American River Parkway.

The City of Sacramento was formed at the location of where the American and Sacramento Rivers joined. Historically the city was subject to flooding when water levels on either the American or the Sacramento Rivers would flood. To protect the city a series of levies was built around the floodplain for the American River. This floodplain area is now the American River Parkway.

The American River Parkway floods seasonally. Is this really a good location to regularize a squattor settlement? Isn't a wise idea to enforce the zoning rules here?

On the otherside of the levy, the land is owned by SMUD, Southern Pacific and the Blue Diamond Almond Growers.

The land owned by the railroad is polluted. It was waiting the approval of a plan to enivormental impact report to clean up this land. Again is this the type of land where squattor settlements should be regularized?

The land owned by SMUD is where the natural gas main to the city enters the community. SMUD has plans for a power substation to be built on the land to support growth on other land owned by the railroad once that land is cleaned up. Before this land was owned by SMUD it was owned by the railroad. It still has detectible levels of PCB's in it. The enviromental impact of the clean up allowed industrial uses here because it was argued that by capping the dirt with a powerstation, the pollution wouldn't spread. Residential uses were prohibited here because the PCB levels are still considered unsafe.

Again is this an area where squattor settlements should be regularized?

Lastly there is the land owned by the Blue Diamond Almond growers. Its the rail spur to serve the factory. Again is this a good place for people to be living? To provide water and sewage to this area, you would have to dig through the polluted soil on the ajoining lots. Its an industrial area. Again is this a good place to regularize squattor settlements?

Its expensive to do an enviromental impact report. Its expensive to provide infrastructure to a development (water, sewage, gas, electricity, drainage etc). In short its expensive to provide people with a legitimate place to live.

The issue is how do we allocate those costs? The planning process allocates those costs onto the developers wanting to develop there properties.

If you think your land would be more valuable as residential than industrial uses, it falls upon that developer of the land to pay for all of the costs of converting that land from industrial to residential uses.

If we didn't the railroad in Sacramento could get away with dumping the cost of cleaning up their polluted land on the public. Is that a good idea? Wouldn't that just be encouraging the railroad to be indifferent to polluting their land?

But once again you are missing the point. The squatters are not squatting private land, as the land owners do what is necessary to protect their property. Squatter camps typically appear on public land or land with no clear ownership, which is what happened in South America, and in this case granting clear title to the squatters drastically improved their material condition.

None of this matters, however, if the planning system that drives people to become squatters is abolished and replaced with a planning system that gives people a legitimate place to live before they have to become squatters.

Clear title is the means of establishing legal ownership. Legal ownership includes a bundle of rights. First it means the state won't take your property without paying for it. Second it means that the state will protect your claim to the land from others encroaching upon it. This is why legal title is such a valuable asset.

The reason the state bulldozes squatter settlements is to protect the title of the legal owners of that land. If someone comes upon your land and builds something upon it. You can have them removed with a motion for trespass. Because you know the state will protect you from other people trying to either take your land or from either adding to taking things away from your land, you can now feel safe in making improvements to your land like building stuff on the land, practicing agriculture on the land etc.

If you don't have clear title to a property a bank won't issue you a home loan to purchase that property. The bank doesn't want to lend money to you to make improvements in the land only for someone else to take either the land or the improvements away from you.

This is why no matter what your credit rating a bank won't make a loan to buy or build a home if you don't have legal title to that land. This is why your assertion that ONLY creditworthiness grants credit is factually untrue.

I can see a very good reason to not regularize squattor settlements. Specifically the state has an interest in promoting clear title. If land owners and bankers know that the state will regularize encroachments upon land owned by others both property owners and lenders have less reason to feel confident in investing and improving property.

Why establish a farm if you know squattors can move onto your land occupy it, build structures upon it and take the land away from you? Why lend money to a farmer when you don't know if he will be able to actually grow a crop on the farm or if after planting squatters will move in build structures upon his farm and take his land away from him? Why build a home on property if you don't know if someone else will move in and take it away from you? Why lend money to someone to build a house if you don't know if someone else will move in and take the home away from them?

How much credit will be available in a society where title isn't respected? Why would anyone bother to make any improvements in land where title won't be respected?

The poor in America are now building squatter settlements, so it's clearly not true that clear title grants credit. Ultimately only creditworthiness grants credit.

You may criticize the flaws of living in squatter towns, but you can't deny the failure of the system that forces people to live in them. I have yet to see a good reason why spontaneous towns cannot be given the same legal title as banker-financed development, so that they too could invest in their property. (This has been used to great results in South America.) It would seem to fall under the principle of equal protection under the law.

When the police come to tear down these communities they make sure the occupants are out of the homes before they bulldoze them. When a natural disaster occurs like a fire or a flood there are no such assurances that occupants wouldn't be in the building. Thus the difference is between life and death.

One of the things a planned city can provide is clear legal title. Part of the reason the housing quality is so poor in squattor settlements is no legal title. That means the occupants don't have access to the credit markets. Thus they can't borrow money to finance the cost of constructing nor improving their buildings. The lack of access to finance limits there ability to use high quality building materials and puts high quality construction techniques beyond their reach.

You claim that planners do not have the power to create growth but that simply is not true. The power to create clear title is the power to create create wealth and fund growth. Clear title is what gives bankers and financers the confidence to invest money in the property.

In the US, clear title means access to credit. This is why a man with no money down can buy a home or an automobile. Access to credit is what distinquishes the poor in the US from the poor in Mexico.

You still haven't really addressed the issue of how to keep people from building in unsafe areas without some type of planning process. You haven't discussed how to ensure the provision of infrastructure to these homes in the squatter settlements. How to provide water pressure in the fire hydrants and how to ensure than fire trucks can move freely in these communities? Nor have you discussed how to pay for this infrastructure. In the squattor settlements how do you fund schools, police of fire protection?

The problem, and I think you are close to realizing it yourself, is that planners were given the power to stop growth but not the power to create growth. So what can they do when the impoverished masses find an escape at the fringe? They demolish it, because that's the only power they have. And what is the difference, when you are a destitute man, between losing everything to a natural disaster like a flood and losing everything to a police bulldozer? You have lost everything in both situations.

The power to create growth is not to destroy and imprison people who build in flood plains, it is to create new neighborhoods with complete infrastructure before people are forced by destitution to build in flood plains. It is ensuring that people always have a real, viable choice. Planners have not been keeping up with demand for new cities, instead they have suppressed that demand, and this is what I mean when I say that the planning system creates a permanent crisis.

What about infrastructure issues? Tijuana has lots of squatter settlements, but the history of them has been pretty bad. Prewar, the squatter settlements were built up in the Tijuana River flood plain. When the floods came, the residents were wiped out. Afterwards the city went in bulldozed the remains of the communities and built a flood channel. Then the squatter settlements went up on the hills. When it rains those areas flood, have mud slides and hundreds of people die.

A big part of the reason city planners were given so much power was to avoid these types of issues. By zoning communities, you could ensure that people wouldn't be building in flood plains or on hillsides with lots of clay in the soil with poor drainage.

With planning you have an idea of how intense land use will be in a given area. That means you have an idea of how much infrastructure needs to be provided and where and how to provide it.

One of the things city planning standards achieve is the prevent building homes too close to each other to prevent firefighters and fire trucks from reaching the structure.

They ensure that homes are built attached to the sewage infrastructure, that potable water is provided and that homes are build in areas with adequate drainage. They even ensure that there are adequate numbers of fire mains with adequate pressure in the water lines to get water to all of the areas that might burn.

Building codes also give you a good idea of what materials are being used in the homes. Bad wiring can burn down not just your own home, but the adjacent homes in a squatter settlement. Building codes mandate the use of flame resistant materials.

If cities are unplanned. How do you ensure that these other aspects of urban planning are addressed in your squatter communities?

Whoever brings the money decides the morphology. If development is paid for by central bankers and mortgage corporations, then necessarily they will decide what kind of houses developers will produce. In a squatter settlement however individuals bring their own capital, sometimes in materials instead of money (collecting scrap from which to build their shelters). This is why the morphology of squatter settlements is so individualized.

So how do you use planning rules to plan for development to be unplanned? If have often wondered whether restricting building and land ownership mostly to those who will occupy the unit would do it. No developers, only builders.

One of the problems with planning is that it artificially affects land prices. Developers make as much (I have no idea whether this is true) from changing the zoning of the land they own as from the added value of building things. Having the value of land increase slowly in the hands of individual homeowners as a result of a lot of individual actions is not as efficient a use of capital as having a lot of land appreciate all at once in the hands of one developer as a result of an agreement between the planner and the developer.

I'm not suggesting anything dishonest about this agreement, although the system carries that risk, simply that planning itself will tend to promote sameness even when that's not one of its objectives.

I think you misunderstood me.

What is wrong with planning via squatters communities? I would think you would approve of the idea of people moving into an area building something that is livable for them and then just regularizing that choice by providing sewage, water, electricity and title to the land after these folks have made the community livable.

If not by squatting aren't you stuck with the developer/city planner system of city formation that you seem to rail against in this blog?

In its most basic definition a market is just an exchange of property between two parties. So to say that developers are "the market" is missing half the picture. If you follow where the property is going, you find out that developers have two markets. Firstly they transfer all the infrastructure and all the upkeep of their new neighborhood onto the municipality, and secondly they transfer the buildings they produce onto the banks, which then rent them out for a period of years until which point the renters have paid enough and they become owners, or the bank forecloses and starts over.

So yes the market is responsible, but the municipalities are the market. They are owning these new neighborhoods. If they refused to do so, the banks would not buy the houses, and the developers would not build them. Their market are the municipalities.

Now what happens in the squatter settlement? The squatters are building to own, or at least own as far as the possibility of a police crackdown allows them to. And who owns the neighborhood as a whole? That's not clear, and it's why the squatters can move in.

It won't help to shift all the blame on developer greed or government regulation. They are two mutually reinforcing aspects of the same modern planning system.

Okay, I agree with you that the city (little c, meaning all the people in the city) accepts long-term liability, but the City (big C, meaning local government and those awful planners) doesn't. They don't own the houses, and they have little control over the form of them. The city buys them, so the city (also often called the market) controls their form.

So, back to my original question. If the developer and the market decide the form of the neighborhood, and the city participates passively (except zoning, but many, many planners oppose single-use zoning), why are you blaming planners. It seems to me that the market is the reason that most of the American built-environment of the last 75 years is inhumane and poorly formed. Now prove me otherwise.

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.

Subscribe to Conversation