Thread by @mathieuhelie

My next paper topic: urban planning for succession or for immutability?

New Urbanism arose from this question: what would be better than a subdivision or a mall? A more complex subdivision/mall hybrid, based on traditional building codes. By hybridizing, it increased the financial and legal burden of the developer, and saw limited success.

The new urbanist project saw car-dependency as a design problem to be solved locally, without considering that subdivisions and shopping malls may be optimal solutions for the context within which they were built.

If we consider that car dependency is an issue of scale, determined by the scale of the regional road network vs the scale of the regional mass transit network, building high density neighborhoods will have the opposite of its intent: create more car-dependent households.

Given those conditions the best course of actions to limit car-dependency is the least intuitive one: the lowest density of new development. If possible, no new development at all. This is now the semi-official policy of the state of California and its cities.

New Urbanism sought to imitate the outcome of traditional cities while shortcutting the process they were built from, like trying to plant an old-growth forest in a desert. This ironically succeeded best in a context that had no traditional precedent: resort towns.

The main concept behind ecological succession is that an ecosystem complexifies when simple organisms complete their lifecycles and create the conditions for higher-order organisms to survive. In other words, they form soil. One community follows another.

en.m.wikipedia.org/wiki/Ecologica…

For all of urban history, urbanization was a nonlinear process. Lots filled in as needed over time, in a process some call incremental growth, or organic growth, etc. It was iterative, driven by acute feedback and extreme scarcity. Even the shape of lots was defined over time.

After WWI, just as the petrochemical industry was hitting its golden age, the world embarked on a brand new experiment of linear urbanization. The need to support motorization meant buildings, streets and lots had to be designed, financed and constructed simultaneously.

The phase transition from nonlinear to linear urbanization started with streetcar-oriented suburbs but finished when the automobile was fully democratized. The last living examples of nonlinear urbanization are hence found just before, and they are dramatic to witness.

This is São Paulo’s Paulista Avenue on the day of its inauguration, 1891. Many people showed up to visit a whole lot of nothing.

commons.wikimedia.org/wiki/File:Jule…

Within a few years the place had been settled with spatious mansions with elaborate gardens as the city’s wealthy elites tried to escape the overcrowded city center.

Yet business kept moving southward, and by the 1950s the central business district was relocating to Paulista, particularly the cultural and media industries which took advantage of its high altitude to set up their transmission towers on their already towering buildings.

In the 1970’s Paulista was the undisputed economic heart of the city, and the country with it. That meant the street had to be redesigned to accomodate more cars and more traffic, which would eventually reach apocalyptic levels and stay there.

Then growth kind of peaked. The business district moved on further south. The few remaining mansions became a protected artifact of what had once been the street’s main land use.

netleland.net/hsampa/mansoes…

...which sometimes involves weird marriages.

The successional economy is the entrepreneurial analog to a successional ecology. There are businesses that can only emerge on the remains of defunct predecessors, they are too vulnerable to exist without a carcass to decompose.

The Belgo building was once a luxury department store, then a garment manufacture, now is filled with art galleries, independent restaurants, a yoga studio, various offices.
en.wikipedia.org/wiki/Belgo_Bui…

This was obviously what Jane Jacobs meant when she wrote that neighborhoods need a mix of new and old buildings to be alive. Some businesses cannot survive in new buildings.

The “retail apocalypse” shuttering malls and big box stores is the successional economy analogue to forest fires - extreme fragility linked together in catastrophic failures. But forest fires leave behind nutrients and light, which are analogues for structural capital and demand.

Gentrification is tragically the end steady-state of economic succession, the trees slowly taking back the regenerating land, suffocating uncompetitive economic activities. There is no alternative to gentrification, but new fires will get triggered.

In Stewart Brand’s How Buildings Learn series, Joel Garreau explained Edge City as the CEOs moving their business to shorten their commutes, in contrast to the traditional pattern. Seen in Paulista business following executives IS the traditional pattern.

Back to New Urbanism, and why it was doomed.

New Urbanism’s ideal was modeled as a “transect”, which is a concept borrowed from ecology.

transect.org/transect.html

It models transitions from one community to another as a sequence of distinct geographical layers.

What is missing from the new urbanist transect is its “geological” dimension - how does one layer arise out of the next one? Without removing the structure of the previous zone and starting over, we cannot “upgrade” a zone.

Suburban development proceeds as such - just level the rural structure, subdivide it and sell it back as lots for houses, offices, malls or warehouses. The New Urbanists thought the same could be done for the urban zones, but the financial risks are an order of magnitude higher.

This means besides some limited success making planned towns (“T4 General Urban Zone”) with committed landlords, New Urbanism has failed at its stated objective of ending suburban sprawl.

It tried to shortcut its own model, going from T2 to T4-5 without first going through T3.

That we can show T5 and T6 zones exist geographically, from a historic trajectory that was pre-capitalist and could not take on large-scale risks, means they had to appear through a successional process.

What needs to be added to the transect is the vertical axis of time.

If we were to roll back the clock and model, for instance, lower Manhattan “devolving” back to its suburban and rural origins, we would see that the T5 and T6 patterns are rooted in a suburban pattern of garden houses and townhouses that has nothing in common with today’s suburb.

In fact, old depictions of Manhattan tend to show the ideal T5 zone preserving a lot of its T3 structure.

What this implies is that, much like today’s existing T6 zone has a past that is nothing like today’s T3 zone, today’s T3 zone has a future that is nothing like today’s T6 zone.

This future needs to be invented.

This is where “accessory dwelling units” get their importance.

Behind the two fracture points of modern planning, NIMBYs and gentrification, is one fundamental question:

Should neighborhoods change?

NIMBYs and anti-gentrification activists agree that they should not. The modern planning system was invented to enforce that agreement.

Introducing change into such a system is to work against its nature. Whether the rules are coming from the national government, the municipal bylaws or property owners’ associations does not matter. What matters is the intent of the rule, to keep things settled as they are.

Hence we have witnessed a massive conflict emerge over the euphemistic accessory dwelling unit. In jurisdiction after jurisdiction, (accessorydwellings.org/adu-regulation…) the pressure of real-estate demand is splitting the constituents: is this what we want in our backyard to keep rents low?

But if you are wondering whether a struggle to add a few permissions allowing property owners to build studio rentals on their properties is worth the pain, realize what this change implies; it shifts the fundamental question of planning from should neighborhoods change to how.

This is not a simple addition of studio rentals but a generational shift in neighborhood planning. When the next generation finds itself occupying a neighborhood filled with studio apartments, there won’t be a need to shift question again, only to provide new answers.

What triggers succession in natural systems? So far succession has been phenomenological, we see it happening but don’t understand how or why. There is an emerging theory around fungal networks that promises to integrate ecology and network science.

Forest soils have been shown to consist of extensive networks of “mychorrizhal” fungi that communicate nutrients and threats from tree to tree, mushrooms being the “fruits” of these underground trees. A single individual fungi can span square kilometres.

https://twitter.com/mathieuhelie/status/1091103859911606273?s=21

Compost researchers such as Dr Elaine Ingham are teaching that the successional habitat of plants is related to their mycorrhizal relationships. Weeds bloom in the absence of fungi, when bacteria outweigh fungi 10:1, trees grow when fungi outweighs bacteria 5:1 up to 1000:1.

The ecological role of weeds is to grow roots that provide an anchor for fungal networks to bootstrap, while increasingly interdependent plant species colonize those networks and grow the land into a underground of increasingly dense and complex communication networks.

The urban equivalent of an old-growth conifer forest is the skyscraper district. The extreme concentration of daytime residents needed for them to emerge requires multiple overlapping mass transit lines. Manhattan famously has two such districts, downtown and midtown.

This highlights the chicken-or-egg-first problem of building transit for districts or districts for transit, which is at the core of urban sustainability objectives.

We need energy-efficient transit, but we’re stuck with weedy urban growth that can’t sustain it.

This is how nature solves the problem: short lifecycle plants are less dependent on networks than long lifecycle. Long lifecycle plants feed network growth more, until the climax ecology is achieved.

The problem is our urban planning has no concept of lifecycle.

Some cities, confronted with the extreme devastation of their cores by decades of policies expecting immutability, are embracing weedy urbanisation, figuring that even container shops with portable toilets are preferable to a cratered block. It’s alive.

granolashotgun.com/2019/03/04/mar…

What does planning that integrates lifecycle look like? It begins from the assumption that everything must eventually be replaced, and therefore the question it asks first is what should it be replaced with?

Where it diverges from linear planning is the assumption of order in the rate of replacement. We do not know precisely when each part must be replaced, but we know the half-life of groups of parts. The most reliable plan thus plans for a random half of the system to be replaced.

To not do this is to invite randomness as a pretext to prevent change. We must rebuild this block, street or bridge exactly as before because it has become a maintenance emergency. If a plan already exists to upgrade it when its lifecycle ends, it becomes maintenance opportunity.

Planning for a half-life has interesting implications. For central business districts, it means facing the fact that large-scale noisy reconstruction of towers becomes a permanent feature to be mitigated for life to function normally on a day-to-day basis.

Lifecycle is determined as much by engineering as by finance. At the other end of the scale, it means approaching new neighborhoods with the assumption that some lots may never be built on, some blocks never filled in, if demand or financing evaporates.

And finally lifecycle means the rate of growth of the city expresses itself through planned or unplanned changes. What if the end of a house’s lifespan means city-sponsored demolition, as happens in the “rust belt”? Or growth pressure is pulling it up a level or two of transect?

Coming full circle, it means cities need a plan for each neighborhood to go up, down, or level in their transect zone, instead of the assumption that the transect zone itself is the plan and must be encoded and enforced. And those plans must work partially and randomly applied.

@threadreaderapp unroll