JULY 29, 2014
Publications > Table of Contents > The Spatial History Project
Piedmont-Palladino, Susan, editor. (2011) Intelligent Cities. Washington, DC: National Building Museum, pp. 36-41. 
The Spatial History Project

Intelligent Cities: Case study
Erik Steiner 1
1. Creative Director, Spatial History Project
This essay was originally published as a case study in the following book: Piedmont-Palladino, Susan, editor. (2011) Intelligent Cities. Washington, DC: National Building Museum, pp. 36-41. 

A project of the National Building Museum, the Intelligent Cities Initiative's mission is "to engage thought leaders, government officials, and the public to think about how to use existing and emerging technologies to improve our built environment." Steiner served on the Advisory Board for the Initiative in 2010-11.

The Spatial History Project at Stanford University brings together scholars who share an interest in using visual and spatial methods to study the past. We consider our work to be part of a larger "spatial turn" in the social sciences and humanities—not a revolutionary change but a convergence of a set of methods and ways of thinking that may offer new insights into the human experience. In our conceptual focus on space, we seek to understand how spatial relationships create history and how history, in turn, generates and informs relationships in space.
What do we mean by spatial relationships? As geographers have long recognized, understanding the location, proximity, and distribution of the earth’s resources is important to understand how the natural and human-made world functions. Equally, the geography of the interactions and relationships of humans can explain a lot about our social, economic, and political behaviors, and the impact those have on the environment. Historians are also interested in how society functions, but typically explain it through a chronological context of what, when, why, and how things happen, but rarely where. Our argument is that the spatial context matters in all of these cases.
We seek to understand history looking through the same lens in two directions. That is, how spatial relations stimulate cultural, social, and political change, and how changes in technology, economy, and policy create new spatial relations. For example, we ask how did nuisance laws prohibiting animal slaughter from downtown San Francisco change the practice of selling and consuming meat, and the distribution of its polluting effects? Or, how did/do changing transportation technologies alter travel times and affect the relative proximity of our cities or spaces within our cities? In the context of this book on the present and future of the city, our work puts value on thinking critically about the complex process of change, and offers methods for contemplating the sometimes invisible connections between spatially distant parts of a modern city.
Spatial history, in this guise, seeks to reconstruct past landscapes in order to understand how they were produced, experienced, and ultimately transformed. This approach relies on a certain expertise in software but otherwise offers straightforward ways of translating historical documents into datasets that can be digitally manipulated and recombined. Our emphasis on space draws us to documents that describe geography but we rapidly move beyond these to other types of data found in the archives that many historians might find unusable. These sometimes obscure and frequently overlooked sources, such as railroad freight rate tables, hospital death records, or even nineteenth century Brazilian novels provide material to reexamine old cities in new ways. In the following examples, I’ll attempt to describe how and why we do this.
In our collaboration with James Tice, we digitally restored a series of detailed historical maps and views of Rome, including the famous 1748 great plan of the city by Giambattista Nolli. Using satellite imagery, we georeferenced the maps in their modern context, and digitally annotated their architectural, cultural, natural, and urban design features (Tice et al. 20051, 20082). The resulting database provides dozens of expansive layers of data that can be combined, compared and brought to bear on a series of historical questions. The interactive maps we have built with these layers fundamentally changes the way Tice teaches courses in architectural design, and ultimately how individual building artifacts are understood. By providing a spatial context to St. Peter’s basilica, for example, you can see that its vast piazza was designed to create a dramatic contrast to the small streets experienced by approaching pedestrians, revealing the impressive ensemble only at the last possible moment as they entered the piazza. Only then do you realize how the 20th century grand boulevard by Mussolini diluted this spatial concept and weakened the intended urban experience (see Figure 1).
Figure 1: Reproduction of the Nolli Map of Rome at St. Peter’s Square.
Note the distinction between the historical and present day approaches to St. Peter’s Basilica; in the original design the square provided a dramatic contrast to the narrow approaching streets.

An interactive version of this tool is available at
Reconstructions of the past not only afford a view into what the landscape looked like at a certain moment in time, but also force us to ask how they were created, and what role they played in shaping historical events. Historical cities were no less defined by complex and interconnected processes than their modern counterparts are today. They were dynamic, constantly evolving places that were built by a series of physical actions but given meaning by the ideas and beliefs of those that interacted with/in them. Spaces of a city today do not merely exist as they are, but have been produced by the way that people use them, talk and think about them.
In this way, places can be described not only by their physical features, absolute position in GPS coordinates, or dimensions in meters, but also by their relational position in hierarchies and networks. In Richard White’s Shaping the West project we have discovered how 19th century railroad companies sought to maximize profit by manipulating the cost of shipping goods in California depending on destination city, and the product being shipped. The effect was to create a geography of commerce that looked nothing like the actual space of California. Some places (those where the railroads had competition from water routes) had cheap service and were thus pulled closer to San Francisco, while others (where the railroads had a monopoly) were pushed to the margins of the state. The railroads could manipulate rates and thus space to bring wheat producing stations closer to San Francisco Bay when they shipped wheat and further away from the bay when they imported consumer goods. These new, fabricated spatial relations in turn encouraged a specific distribution of agricultural and manufactured goods production that defined the historical trajectory for hundreds of communities across California (White 20113)(see Figure 2).
Figure 2: The Effect of Railroad Freight Rates on the Geography of California, 1876
This figure provides a graphical demonstration of how railroads constructed and manipulated space, pushing and pulling stations in relation to San Francisco depending on the goods shipped. Farmers and merchants were at the mercy of the railroad officials who had vast power over their economic well-being.
At the scale of a city, similar historical processes shape public spaces of town squares, theatres, and churches, and the private spaces of residences and professional offices. In Zephyr Frank’s Terrain of History project we explore how spatial patterns of disease, property values, and occupations have uneven and unequal effects on neighborhoods and the people of nineteenth century Rio de Janeiro. Economic changes in the city in the late nineteenth century meant changing forms of commerce and a massive influx of immigrants, creating new connections and breaking old ones in the city. This urban reconfiguration resulted in changes in how the spaces of the city were experienced, for example, by artisans who were displaced from their traditional cohesive neighborhood and network into smaller and less-connected clusters as wealthier businesses moved in (Frank 20074). Or, by slaves who could be bought and sold to owners in different parts of the city, and in one day face dramatic changes in their socio-economic circumstances (Frank and Berry 20105). By tracing and explaining thousands of these flows and transitions, layered one upon another, we hope to begin to decipher their complex networks and processes of interaction. It is in this continual space making and remaking that a city is created.
But what we understand about the creation of cities is limited to the sources available to us. Many sources (even recent ones) are often incomplete, inaccurate, or not easily adaptable to computational methods, no matter the power of the supercomputer. How can you fully describe a process or movement if all you know is the location of Point A and Point B? In our Geographies of the Holocaust project, we have this challenge in our study of the ghettos in 1944 Budapest (Giordano and Cole 20116). We know where Jewish residents lived, and where market halls were located. And, we know that Jewish residents were only allowed to leave their homes from 2-5 pm to do their shopping. But, what effect did this restriction have on the street life of the city? What new spaces were animated that were previously inert? By creating a geographic information systems (GIS) model and simulating the hypothetical journeys of thousands of Jewish residents from home to market hall, we can begin to answer these questions. While the precise paths, walking speeds, and shopping time are all uncertain, by animating thousands of potential journeys, a picture quickly emerges to reveal the important role of specific streets and intersections. Some areas become choked with activity, others remain empty (see Figure 3). Places of gathering perhaps served to reinforce a meaningful although temporary sense of community, while in other places (and times) Jews would find themselves alone and vulnerable under the watchful eyes of the Hungarian police. The ’doctors of space’ that created this new geography sought to reduce the visibility of "the Jewish problem" in wealthy areas of the city, and concentrate it in others; ultimately, our analysis suggests that Jews (marked with a yellow star) were still relatively visible on streets that had been cleared of Jewish residences.
Figure 3. Mapping Mobility in the Budapest Ghetto
A screen capture of a digital animation showing Jews walking to market halls from the ghetto in Budapest, 1944. The requirement that Jews only shop between 2pm and 5pm made the Jewish population concentrated in some areas while almost completely invisible in others.
Simulations and models can help account for uncertainty, and are powerful ways to mimic dynamic and functioning systems that can then be observed and analyzed. But what if we could measure and observe past societies more directly, with an appreciation for their complexity but with fewer of the assumptions we bring from our modern context? Zephyr Frank’s research explores the value of nineteenth century Brazilian novels in describing the past in such holistic and organic ways, informing our understanding through rich personal histories and trajectories playing out in the context of a historical city. In much the same way that Jane Jacobs showed us the view from the street and how it explained the function of cities, Frank argues, novels reveal the city in all its messiness, while still grounding itself in plausible fictional characters situated in real places. While the spaces they represent may be chaotic, novels need to make sense of it all (to be an enjoyable read, at least). In this way, great authors reveal ways to see and synthesize an experience of a place better than any of our simulations and models could ever do empirically. Frank’s exploratory methods, such as network visualizations of characters, scenes, and locations, reveal at the very least how certain places of a city become meaningful to people over time.
I hope from the preceding examples you can more clearly understand how our work can shed new perspectives on cities both past and present. I conclude with a few comments on the specific methods we use in our work. We claim no novelty our tools, but their use in each of the contexts described above is original.
The language of space and place well-known to geographers, architects, and urban planners is not new to historians, but the tools and methods are only recently gaining acceptance and widespread use. Our projects are distinct from traditional historical practice in that they are strengthened less by narratives and more by visualizations, and are dependent on the intelligent use of computer software such as Geographic Information Systems (GIS) (White 20107). The use of GIS allows us to reconstruct and analyze our spaces, and visualization tools can animate the movement and change within them. These tools are commonly used by professional analysts for such things as determining the effectiveness of a public transit network, identifying optimal site locations for a new retail stores, or measuring the impact of climate change on vegetation in high alpine areas. And, they are used implicitly by each of us on a daily basis to choose driving routes, identify if a bus is running late, or search for nearby restaurants or local news. Most importantly for our work, GIS can quickly process large amounts of raw coordinate data to determine statistically significant hotspots, anomalies, and general geographic trends.
GIS is one of many ways of organizing data about the past, and it does a decent job of it. We are always careful to recognize, however, that these computational (and often convincing) methods can ignore and distort as much as they reveal – fixing historically-produced spaces into absolute positions and privileging raw distance rather than meaningful connections. Ultimately, the power of amassing massive geographic datasets is limited less by the capacity of technologies to collect and process them, and more by our human ability to synthesize and understand their meaning. In our work we use visualization as one way to address this challenge. Visualization allows us to explore, understand, and eventually communicate about the complex patterns that large datasets reveal. Drawing on the work of geographers and computer scientists who have developed visual and interactive methods for exploring dynamic spatial data, we aim to create visualizations more often to help generate new ideas and hypotheses than to represent final research results. They are a way to identify a set of questions, prior to answering them.
In our lab, we have discovered the power of data to recreate past landscapes – both real and imagined. Our collaborative projects span a variety of scales and historical events, and are continually challenged by the difficulty of interpreting data of uncertain completeness, quality, and provenance. Not only a powerful way to make new arguments or substantiate old ones, visualizing history is a means of generating many more questions that would otherwise go unasked, and telling stories that would otherwise go untold. In the end, these stories often teach us that the world is more complex than we initially suspect, and that we should proceed with mindfulness in the care and design of our future cities.
End Notes

1Tice, James T., Erik Steiner, and Allan Ceen. Interactive Nolli Map Website (2005)

2Tice, James T., Erik Steiner, Dennis Beyer, and Allan Ceen. Giuseppe Vasi's Grand Tour of Rome (2008)

3White, Richard. Railroaded: The Transcontinentals and the Making of Modern America (New York, 2011).

4Frank, Zephyr. "Layers, Flows and Intersections: Jeronymo José de Mello and Artisan Life in Rio de Janeiro, 1840s-1880s." Journal of Social History (winter 2007): 319.

5Frank, Zephyr and Whitney Berry. "The Slave Market in Rio de Janeiro circa 1869: Context, Movement and Social Experience," Journal of Latin American Geography, 9 (2010): 85-111.

6Giordano, Alberto and Tim Cole. "On Place and Space: Calculating Social and Spatial Networks in the Budapest Ghetto", Transactions in GIS, p. 143, vol. 15(s1), (2011).

7White, Richard. "What is Spatial History," web publication, link, (2010)