Cities as systems of encounter

The Sociology of Space – by Martina Löw
November 26, 2016
Cities as systems of communication
January 13, 2017

Encounters in the streets. Source: Architecture and the Ballet of the Street, short film (Netto et al, 2015)


All organisations involve the co-ordination of interaction in flows of time-space relations ‘channelled’ through regularized contexts and locales.

Giddens (1984:77)


The recognition that the ‘encounter’ is an important fact of social life is not new, of course. It was asserted by the work of sociologist Erving Goffman in the 1950s and 1960s. But it was in the works of Anthony Giddens, Bill Hillier and Julienne Hanson in the 1980s that the importance of space in the creation of encounters was introduced in social and spatial theories. In the first part of the book The Social Fabric of Cities, I draw from such traditions in order to understand the role of cities in encounters, and how the creation of encounters is one of the major results (and goals?) of cities. I do so through what appears at first sight an unusual doorway to the role of cities in encounters: I look at the conspicuous problem of social segregation.


Most studies of segregation are limited by a somewhat static and territorial view, in which the role of space is still confined to physical separation, and segregation is still seen as the social effect of physical separation. But how can we explain the experience of segregation when socially different people actively move around the city? Making use of ideas ranging from social networks theorist Linton Freeman’s to the time geography of Torsten Hagerstränd’s, the book argues that segregation manifests itself silently in the form of restricting contact – a subtle reduction in the potential of the encounter, and a form of social distancing that comes into being through the very trajectories of different groups in a city. And this is a much more elusive form of segregation than that of residential territories.


We can explore the mobilities of actors shaping the delicate conditions for a temporal ‘geography of encounters’, a material account of how encounters might happen. For instance, is mobility an absolute capacity, shared by social actors equally? Different forms of mobility might be associated with different social groups and different forms of urban experience. In order to assess the role played by mobility in the creation of encounters, we can dismantle the complex web of individual actions that make up our daily routines in cities in recognisable networks of movement. I develop this look into segregation as subtle networks of trajectories in urban space in two empirical studies.


Segregation and potential co-presence in Niteroi, Brazil. Source: The Social Fabric of Cities

Some empirical findings


Although such physical separation is undoubtedly a form of segregation, it is segregation of a very limited and special sort. . . . All restrictions on interaction, whether they involve physical space or not, are forms of segregation – in social space.

Freeman (1978:412–413)


My research collaborators and I developed a first empirical study in Niterói, a town in the metropolitan area of Rio de Janeiro, Brazil. We interviewed and mapped the daily paths of 121 people from different income levels, and overlapped these paths to visualise where they move through and their potential places of encounter (figure 1).

Trajectory patterns of different income groups in Niteroi, Brazil. Source: The Social Fabric of Cities

Trajectory patterns of different income groups in Niteroi, Brazil. Source: The Social Fabric of Cities

Different networks converge in the Central Business District (CBD) – vehicular movement along longer streets, and a strong mix in pedestrian movement, with broader superimpositions of the routes of lower income and middle-income actors. Although the study can only show trends in segregation, potential for encounter can be observed even within the trajectories of a limited number of actors. Lower income actors display a higher proportion of exclusive paths, indicating signs of residential segregation, whereas lower and higher-income actors share only 2 per cent, especially near the CBD. This indicates a strong dynamic segregation between the richer and the poorer. We assessed mobility in personal urban trajectories using a composed measure of extent, number of activities performed, and the number of street segments used by actors, also dependent on modes of transport (a high number of segments indicates the use of many spaces). This method interprets mobility as a relation between the fractal dimension of trajectories in urban space and the number of activities performed in time. The analysis of mobility shows strong differences, consistent with income levels: higher-income actors are likely to carry on more activities with less effort and a ‘finer-grain’ appropriation of the street network.


We also analysed sociability and diversity in the personal social networks of interviewees, considering the different situations in which people make social contact. Middle and higher income actors have higher diversity in their social networks than lower income, allowing them a more ‘amplified’ sociability. These results show that mobility is a key factor in diversifying socialisation. They support the idea that mobility shapes opportunities for encounter, increasing the probability that trajectories of socially similar actors will overlap.


Handling larger populations

The next step was to look into segregated movement of a larger population. New information technologies and social media such as Twitter offer tools for precisely those ends. The set of variables provided by Twitter is publicly available through a principle of anonymity and includes user IDs along with timestamp and geographical coordinates for each tweet. Metadata were collected from tweets posted in the city of Rio de Janeiro through the official Twitter streaming for a 56-hour period. Filters were employed to ignore users providing insufficient data. We inferred residential location by matching recurrent tweet positions with income data available from the Census Bureau, leading to a sample of 2,543 users, resembling the income distribution in Rio. Trajectories between tweet positions were recreated through shortest paths as a reliable proxy. The result was a network of thousands of routes within Rio. It turns out that, as in Niterói, the poorer and richer groups share very few spaces – only 0.8 per cent of trajectories in Rio.


Segregated networks: blue (low income), green (lower/middle), yellow (middle), orange (middle/higher) and red (high income) groups. Source: The Social Fabric of Cities

Then we looked into where class networks converge more intensely. We analysed social diversity in the streets through Shannon information entropy. Streets with equal shares of all income groups contain the highest social diversity. Diversity levels were associated with colours from blue to red (figure 3). There is a small network of socially convergent streets with an interesting superimposition of trajectories of users of all income groups in denser, busier areas. These are the most likely spaces for finding socially different people.



Social diversity in the streets: levels of superimposition of trajectories of Twitter users of different income groups, from less (blue) to more intense (red). Source: The Social Fabric of Cities

Considering the role of social difference and mobility in the formation of segregated networks, we found that:


  • Different income groups have different mobility levels. Restriction on mobility affects the capacity for encountering actors in different spheres of sociability, and therefore the capacity for creating more diverse personal networks. In other words, there is potential influence of mobility in the diversification of sociability for different income groups.


  • Similarities in mobility patterns increase the density of encounters between socially similar people (that is, actors sharing similar income levels and probably lifestyles), increasing homophily (the tendency of actors to associate with those with similar characteristics to themselves) in personal networks. Homophily varies according to levels of localism (dependence on proximity to home a dependence on proximity for creating social relationships), as does mobility. We performed a social network analysis of people grouped according to income. Links show the number of potential encounters identified between pairs of income groups (vertices in the graph below).  Encounters are seen as more likely between socially similar people.


Intensity of encounters between different income groups. The shorter and thicker each link is, the higher the number of encounter between income groups.

Intensity of encounters between different income groups. The shorter and thicker each link is, the higher the number of encounter between income groups.

  • Different levels of diversity in personal networks are associated with different income levels. In principle, the more complex the mobility pattern in relation to places in the city, the broader the potential to amplify and diversify personal networks.


  • Differences in mobility patterns disconnect contact opportunities between the socially different, making the formation of personal networks across social classes incompatible.


These findings mark a shift in focus away from the relatively stable segregation of places – where separation is assumed rather than shown – towards the role of space in mediating contacts between different social people. The book explores an understanding of how elusive forms of social difference penetrate everyday life, to become social distance. Cities are forms of dealing with the elusiveness of encounter, this fragile foundation of social life – perhaps as fragile as its passage to actual interaction and communication between people. Cities are crucial for that passage as well.


Read the book’s Contents and Introduction

The book on Routledge and Amazon

Promotional code for 50% discount:

in ‘Purchasing options’ choose hardback and ‘add to cart’

then enter code ASHGATE230


Be Sociable, Share!




  1. Jona says:

    I have been browsing online more than 3 hours today,
    yet I never found any interesting article like yours. It’s pretty worth enough for
    me. Personally, if all website owners and bloggers made good
    content as you did, the net will be much more useful
    than ever before.

  2. Dierci Silveira says:

    Dear Vinicius,
    I found fascinating your blog and post. Your approach can be useful for many other applications in urban settings and its social demands such as energy use.
    For instance when studying Urban Metabolism and Energy Consumption your approach can drive us to possibilities to tackle public energy policies at residential and commercial buildings.

Leave a Reply

Your email address will not be published. Required fields are marked *