Image from Creative Commons, Photopin.com
In the previous installment of our City Science series, we speculated on the future of an Urban Science approach to city design and the importance of articulating values in advance of parametric solution-seeking.
One central debate in urban design and planning is does form follow function? First introduced in its most palatable form by Louis Sullivan and advocated by the early Modernists, this principle suggests that an object’s physical shape is derived from its purpose. In the next two chapters of CoLab Radio’s City Science series, we’ll take a look at how City Science reshapes our thinking of urban form and function particularly in two domains: spatial planning and energy.
Spatial planning engages several sectors of society. Arguably, it is the method by which the very character of the city is designed through lenses of transportation, land use, environmental and community development. Early intuitions that these parameters could be programmed into a rational framework failed to respect or even consider humanitarian ethics (see Can Cities Compute?). This thread of critique continues to bind today’s urban science initiatives. However, greater computational power afforded by today’s technology and a growing emphasis on interdisciplinary thinking may address these important concerns. Such advances help us better understand the systemic causes of certain social conditions and identify solution-spaces through the recognition of distinct patterns across all domains of spatial planning.
City Science “thinking” reshapes our understanding of spatial planning by quantifying relationships. Rather than observing causality alone, urban scientists armed with high-resolution data may be able to find persistent correlations between and within the social, economic and physical elements of the built environment. The idea is that by statistically teasing out relationships between factors invisible to the naked eye we can make robust improvements to urban systems. Rather than evaluating design interventions through cause and effect, we can explore how the effects of these changes ripple throughout a dynamic system—how the urban fabric which is defined by relationships beyond the calculation of any one human mind, might respond to targeted change.
This is predicated of course, on the notion that cites are more than the sum of their parts. Which leads us to contemplate once again, is the boundary of a city necessarily physical? Do we integrate all the demographic data, live twitter feeds, transportation or crime patterns into our understanding of what a city is? If so, then spatial planning becomes an exercise in complexity– that is, we must know how relationships between these factors at small scales generate emergent, observed patterns that characterize our experience of today’s cities.
It takes one to know one. Scientists of physical, chemical and biological systems treat their subjects as “complex” systems. Complexity in these cases doesn’t mean that a system is more obtuse or difficult to decipher. Instead it refers to a structural pattern that emerges from a set of local interactions or interdependencies. Complexity is actually quantifiable (precisely measured in bits), but we’ll keep the philosophical physics for Course 8. The difficulty of defining complex systems lies in the problem of scaling from small local interactions to the larger emergent properties they generate. For example, your eye color is a phenotype that emerges from complex interactions between proteins that are coded by genes read by polymerase on a specific chromosome…and so on. Analogously, specific growth patterns can be observed in informal settlements, but these trends are the results of countless interactions between hundreds of people, the resources available to them, their reaction to the topography of the landscape, cultural building practices etc. These very complex local interdependencies work together to produce an “informal settlement typology” easily observable (just like your eye-color) from an aerial photograph.
Excitingly, some of the emerging tools in Urban Science are also used to study complex systems in the natural sciences. These tools include factor analysis, structural equation modeling, self-organizing maps, cellular automata, network topology and of course, parametric equation modeling. If you’re curious about the interworking of any one of these, particularly in urban contexts, please see the resources provided below. What these modeling tools have in common is that they are all windows into complexity as they attempt to measure relationships between those small-scale interdependencies and the larger phenomena of interest.
Self-Organizing Map visualizing US Congress Voting patterns (visualized using Synapse). Creative Commons, Wikipedia.
Cellular Automata Model for Informal Development. Emily Royall & Amir Mousavi, 2013.
Funny enough, while Urban Science appears to be the cutting edge of adapting complexity-modeling tools to urban contexts today, the concept is quite historical. In a seminal 1951 study, “British Towns: a study of their social and economic differences,” C.A. Moser and Wolf Scott classified towns into distinct clusters based on socio-economic factors. In so doing, they discovered the social and economic underpinnings of what makes an urban fabric industrial, suburban, or anything in between. The trick is that these classifications emerged from the analysis itself, and were not imposed on the data from the outset.
What implication does this have for spatial planning? Clearly, if we can map how our design or policy decision-making will affect distant points of the system, we can make better choices about land use, location of amenities, placement of growth boundaries or TODs and even the nature of zoning restrictions.
Not to say this is an easy task. When overcoming the complexity of the urban data of their study Moser and Scott lamented:
“We find ourselves virtually in the position of the astronomer who looks up at the stars, trying to ascertain not only the distance between them on a plane but also in depth, while the whole universe curves away into a fourth dimension…”
Moser and Scott intuited that cities were like natural systems in their own right, in part because of their wondrous complexity. The City Science approach to urbanity likewise appreciates cities as complex systems, and hesitates to predict successful intervention without first comprehending to the best of its ability, the co-dependences that interact to produce them.
Emily Royall is an MCP student in City Design & Development at MIT DUSP.