In the summer of 2009, I went on a month-long road trip with Jane Jacobs. She accompanied me in the form of her now-classic work The Death and Life of Great American Cities. In its pages she challenged me, as she has so many others, to think more carefully about something we think we already know about: people and the communities they form.
I’d had a lifetime of opportunities to observe a significant number of villages, towns, cities, and landscapes across North America. But during that 2009 road trip I was challenged by the degree and kind of attention Jacobs gave to the invisible social fabric. When present, the social fabric of a city makes it a human habitation, but when it’s thin or absent, city life devolves to the primal level of mere survival. Jacobs observed that the design and function of a city either enriched the human interaction and well-being of its citizens, or it warred against it. She especially lamented how often the latter was occurring in the face of postwar urban development in the beloved cities of her United States.
Jacobs’s astute eye and probing mind extended to every corner and tuned to every dynamic, prompting a series of thoughts on my part: What if we could all see this invisible social fabric of our cities? What if the vision of Jacobs, her remarkable social acuity, were translated into an image that the rest of us could gather around and look at?
Mapping the Invisible
I had, during the eight months prior to that summer, worked in and walked through the downtown core of Hamilton, Ontario, a so-called rust belt city that was fighting back from an economic earthquake in the global steel and manufacturing sector decades earlier. In my “worst of the Census Tracts” neighbourhood, I saw evidence of strong social ties that did not show up in the socio-economic mapping that had drawn public attention that year. It was a neighbourhood with many significant needs, so it wasn’t that the other measures were wrong. They just seemed highly selective and therefore incomplete. I was living in one of the least-favoured Hamilton neighbourhoods for those months, and walking was my only mode of transportation. My own observation tuned by Jacobs’s prescient gaze, I formulated the core question that is still with me: What does the social fabric of this city look like? This neighbourhood? This place?
Over the past nine years, I have been in active pursuit of this social MRI of the city.Creating new images of this deeply complex and largely invisible world of relational ties has been my own Mars mission. Like so many large-scale projects, a simple desire can mask a massive challenge.
In the early months I thought my search was to find out where the social MRI was already being conducted and then make use of the labour of others. The longer I looked, however, the clearer it became that what I was looking for did not yet exist, though many elements were emerging. So I enrolled in a PhD program at the University of Waterloo School of Planning, with the intent of pursuing this question in a disciplined way.
For the past six and a half years, I have been trying to keep the multitudinous forces in and around that pursuit aligned and pointed in a productive direction. I have not always succeeded, and the difficulties of compromise, dissuasion, and indifference are ever present. I’m happy to report there is still a coherent aircraft of sorts driving itself through the clouds—a bricoleur’s craft assembled from a great many philosophic, scientific, and empirical elements.
I do not use the idea of the bricoleur by accident—it is an inspiration from the lovingly cobbled reflections of Stan Dragland in The Bricoleur and His Sentences. The term “bricoleur” in this sense represents the activity of a person who gains what he or she needs by pulling together found items, what is available, rather than waiting until all the “proper” materials are present. Assembling a piece of IKEA furniture is the opposite of being a bricoleur—objects and instructions are complete, ready to be correctly assembled in just one way.
I knew my pursuit would need scientific care, a disciplined, formal means of paying attention so that others could also understand and interpret what I had produced. The bricoleur always faces a tension in such things. If you are the only one who sees the coherence, you will be labelled idiosyncratic and duly penalized by sanction or exclusion. I have experienced both. Robert Louis Stevenson, in his glittering, tragic love story The Toilers of the Sea notes: “Innovations always contend with this difficulty; very few wish them well. The least false step compromises them.” Indeed.
My instinct to stay close to the scientific disciplines was honed through a lifelong respect for the explorers, researchers, pioneers, and risk-takers who preceded me—people who in books, magazines, films, documentaries, and personal contact filled my life with a world of amazing insights and new possibilities. My social-MRI machine would, it turned out, need a phalanx of books, articles, research papers, computational hardware, well-crafted software, specialized instruments, and human participants willing to undergo scrutiny. Over time the effort took shape around three core elements, the anchor pieces of the bricolage.
First I needed to understand the social world of the participants. To that end I developed a new forty-one-question survey instrument called the Social Capital General Social Survey (SCGSS), which covered demographics, social networks, and perceptions of trust—important ingredients without which the idea of the “social” dissolves. I gleaned those elements from a careful review of all the General Social Surveys administered in Canada between 1985 and 2013.
Second, I needed an easy way to measure proxy phenomena for this invisible reality; that is, I needed a way to “see” social capital. There are some very significant challenges in making invisible social ties visible. It is expensive, invasive, and constantly changing. Imagine trying to take a snapshot of all of your relational contacts—online, offline, at all levels. Now imagine trying to get that in a dynamic, movie form. We have rough proxies for it—Facebook networks, LinkedIn, Twitter feeds, geotagged photos, and so on—but a real measure of that social world is even more challenging. My thought was that if you could see a reliable connection between “social” and “spatial,” then you could make some inferences about the former from the latter. Why does that matter? Because spatial data is much easier to collect. Nearly every one of us who uses a smartphone consents to the collection of such data.
A proxy measure is typically a data source that is used to measure a phenomenon that is either difficult or impossible to measure by other means. Think of a black hole. Because by definition we can’t see a black hole (since even light cannot escape its gravitational pull), we can’t send out a black hole ruler or a giant tape measure. They are impossible to directly measure or detect. Instead, we see how light is bent by “something,” and that invisible something turns out to be a black hole.
The third element was to compare the social and spatial patterns to see if they had any relationship with each other. Did people who moved around more frequently or more widely have greater social resources (social networks, higher levels of trust, and so on) than those who moved in a more constrained way?
I assumed a simple logic. If you never left your home, for example, you would likely have fewer social resources than someone who was active in the community. Highly social people don’t tend to sit still, and when they are in motion, they increase the chances of encountering, engaging with, and relating to other people.
Getting Smarter About Cities
Armed with this framework, I set out to capture the data I needed. The first data capture was the social data acquired through the phone lines of randomly recruited participants in a given geographic area. After more than six thousand phone-call attempts, a willing group of ninety-seven was secured and completed all parts of the study—a lot of effort for a final pool. I understand. Who wants to get a random phone call from a research firm asking for your time and attention? I am deeply grateful to those ninety-seven people.
My second data capture was from those same ninety-seven people, who had, as part of agreeing to the whole study, consented to carry a global positioning system (GPS) position recorder with them for seven days. The little device recorded their longitude, latitude, time, elevation, and speed every fifteen seconds. (It could only record where they moved locally on the device—it had no transmit capability.) As part of the project, I developed my own GPS devices (called TRIAT—Tiny Researcher In A Tube) using open-source hardware (Arduino platform based on the Atmega 328) and software (C++) with an eye toward future researchers who may need such devices to pursue more disciplined research than is sometimes available with open, organic data.
Those data points became the basis for creating a time-ordered digital line that could be laid on a map and then analyzed mathematically for a variety of spatial properties: distance travelled, overall direction of travel, total area of their range, and average speed. This was clearly an area where the science-driven nature of the work would come into play as all that data was sorted, cleaned, analyzed, and interpreted—a socio-spatial set of footprints across the city.
During the difficult years of ethics reviews, data collection, comprehensive exams, and a stream of technical details, the thought of the data producing a visual image of the social fabric of a neighbourhood was an impetus to keep going. I felt like Paul Lauterbur, the chemist who puzzled out the early MRI machines and used himself as a subject in pursuit of those early images.
My goal was to see the invisible, to picture the social fabric of our communities. Sticking with the MRI metaphor, you might think of this as moving from an X-ray image to an MRI image. Since we’ve been better at producing “X-rays” of our cities—we can see the roads, pipes, buildings, streets, the bones of a city—planners have given those things top priority. Capital investments are geared to such tangible outcomes.
But we need to see what X-rays cannot—the tissue that holds the city together. City planners and developers have often underserved the soft connectivity of cities—those very complex and dynamic social worlds we are all part of—because it is less visible, less direct, and more complex than the physical city. And because we haven’t had a way to see it. My goal is to change that.
If we want to address issues such as social isolation, we must be able to see the social fabric better—we need a social MRI of our cities that reveals all of the connective intricacies and flows.
A necessary part of attending to the needs of individuals is to attend to the structural dynamics that individuals interact with. This is a core city-design principle. When we see the invisible social structures more clearly, we will know how planning, policy, and development decisions are damaging or nurturing those structures.
The Invisibility of Isolation
When I was first able to see the fruits of my labour—each plotted line representing a unique life—I found the images simply beautiful; elegant and entirely distinct traces of real people moving in and through this world, human contrails made visible by the magic of microelectronic devices linked to a global system of twenty-seven satellites orbiting 12,550 miles above the earth.
What can we know about a seven-day segment of a full human life? What insight can be gained from forty-one questions about profoundly complex inner perceptions like trust of a given individual? Though DNA gives us insight about some otherwise unknowable aspects of an individual, the amount of information we still don’t know about them is many magnitudes of order larger—almost everything about their life, in fact. But the information our knowledge of DNA gives us is nevertheless extremely valuable. The same goes for the numbers I gathered. Insight can be gained through partial knowledge. We don’t need to know everything about a phenomenon to know something valuable.
I still needed to review the SCGSS numbers for missing or errant values, and to analyze them for their conformity to the statistical standards of normal distributions. Numbers must be forced to dance within the confines of a nice bell curve suitable for regression analysis—that is, how changes in one variable cause changes in another variable. These changes are part of the exploratory process. Imagine these computational structures as a machine that can only handle apple-sized fruit, but you have, in addition to apples, a watermelon and a grape. You have to be clever to find a way to fit the watermelon or grape into the apple machine while retaining their unique characteristics.
Respondent-driven surveys such as the SCGSS take highly complex, synthetic psychological states of individuals and convert them into a numeric scale. In my research, this conversion process enables me to tell from one person to another, for example, whether they have more or less trust in something or someone. So in the survey I asked:
How much do you trust people in your family?
And the possible answers were:
[1] No trust; [2] Low trust; [3] Some trust; [4] High trust; [5] Complete trust; [999] Prefer not to answer.”
Converting complex phenomena in this way is far from innocent and can have significant margins of error, but it is necessary to gain even partial insight.
The spatial data is different. It is a behavioural measure—empirical data about where participants actually travel over the course of a normal week. It must also be interpreted, but it required a different set of analytic processes.
I had to convert the 2.1 million GPS-recorded geographic points into a form analyzable by spatial software. These pathways had to be processed into spatial measures using other machine processes. When you have more than twenty-five thousand observations on five different variables for each of ninety-seven people and must perform complex calculations on their features, machine processes become essential.
The GPS lines were processed using a geographic information systems (GIS) platform, which represents many lifetimes of investment and attention to develop. I made use of that existing software and designed computational models that used various GIS tools to sort, classify, plot, calculate, and synthesize the many points according to the ninety-seven unique lives on those ninety-seven lines. For example, it turns out that we move around our world, pursuing our lives, but with almost no random signature—that is, the individual human paths were not random. Others have noted this.
A University of Waterloo School of Architecture graduate student, Geoffrey Christou, decided that for part of his master’s thesis, he would take a random walk through a city. The experience was, he says, telling:
To attempt an answer to the question of randomness in our lives, I used two strategies: the first was an empirical test in which I went on a random walk, while the second was a quantitative distributions analysis of GPS data points. For the random walk I used a smartphone application to randomly select which street to take every time I came to an intersection in the city of Rome, while carrying a GPS, and making notes. I set out to complete two hours of random walking. This proved impossible. The experience was traumatic, and I do not recommend it to anyone. From this experiment I learned quite definitively that I do not move randomly. The amount of psychological “pressure” which built up after only a few dozen random decisions was overwhelming, and I began to enter a state of extreme anxiety, verging on panic. Those who assert the operation of randomness in the lives of others should perhaps first test this idea on themselves.
We don’t all move around the city the same way—at least not in terms of where we go. Our commonalities, however, are many. The modes of movement are common in their ratios. For example, car in city (“City”), car on highway (“Highway”), walking (“Walk”), staying in one place (“None”): these patterns were much the same regardless of income. Amid the commonalities, however, were some significant individual differences in belonging, number of friends, connections to neighbours, and so on. This suggests that people experiencing social isolation may, in many other regards, lead outwardly normal lives. Isolation is harder to see than we might have realized.
I do not know what will become of this bricoleur’s work in the hands of my doctoral committee and others who have a vested interest. What has become clear, however, is that without improved resolution and understanding, many of the intricately woven challenges we face stand very little chance of being addressed and overcome.
Cutting blindly into a patient and hoping a procedure will work may have some slight probability of success, but it is an approach to healing that few would submit to willingly. That we have often done so in our approach to urban development and city building has been no less tragic. When I think of my mentors such as Jane Jacobs, Ivan Illich, or Christopher Alexander, I am worried that attempts to carry out social-isolation interventions, by governments, agencies, or well-meaning not-for-profits, may well make the situation worse.
Sharper insight is, in itself, no guarantee of better action in attending to matters such as social isolation. Without better understanding, however, we can never hope to be any more than lucky in our most well-intentioned efforts to make the city better for human life.
