Policy Briefing: A smart city’s perspective

Policy Briefing: A smart city’s perspective

Emma Uprichard (University of Warwick)

Cities are, now, according to some, ‘smart’ and apparently, they are going to get smarter and smarter. Since more and more people will be living in cities, smart cities are a way of concentrating efforts, targeting resources and scaling up networks, they are where big data and urban regions come together to create sparkly data fairylands. So if you thought the craze about big data was absurd, then you have seen nothing yet. Cities are the spaces where the big data magic gets sucked up into a series of transport grids, traffic updates, colourful dashboards monitoring energy, pollution, wind speeds, and anything else that can be measured and ordered, and where mobile devices track and trace where, when, how we move. Smart cities are where both the little and large data collectives are watching carefully to cash in on the new social shapes that are being wired up in real time and where virtual sniffer dogs are prowling to detect the ‘baddies-to-be’. In effect, smart cities are the spaces that track human behaviour through huge seismic swathes of dynamic spatial-temporal big data in ways that have yet to fully realised.

For all the hype and hope surrounding big data in cities, smart cities are still also just cities. The idea that cities are ‘more than’ physical geographic spaces is as old as the urban itself. Whilst big data may be transforming cities, the capacity of big data to change social life depends on some of the a priori features of urban space itself.

Smart cities, then, have all the properties of more established notions of the urban in spite of their relative novelty with respect to the big data issues they bring. Therefore, one way of approaching them is to ask long established questions about cities in general. One such question is raised by Beauregard (1995:59), who asks: ‘If the city could speak, what would it say to us?’ In this short piece, I take this question and imagine what smart cities would tell us if they could speak (in doing so, I also echo, albeit playfully, the growing anthropomorphism implied in the idea that cities (and other objects) are now ‘smart’ because big data driven entities supposedly gain a form of ‘agency’). Overall, this perspective takes on the politics of data specifically from the perspective of smart cities and highlights three main points that smart cities might make if they could ‘speak for themselves’.

The City Speaks:

1. We cannot be known using big data alone. Big data cannot describe or explain or predict us. In fact, you will never be able to control us. We are complex. We emerge from, and consist of, multiple, interacting, multi-level nested dynamic interactions – and because we are fundamentally nonlinear systems, we will self-organise, re-adapt and re-describe ourselves every time you try to control us – and even when you don’t. If you want us to change, we need to have a conversation about what we keep the same. If you look closely at where and how we grow, you would see what Kauffman (2000) has suggested, which is that we need parts of us not to change in order for us to change. It’s just how we are. Just as it is difficult to kill or destroy us, it is also difficult to transform us completely. A lot of people have been saying this on our behalf for years now.

To be fair, some smart cities don’t want to own up to the fact that big data isn’t going to help us to radically change, assuming we wanted to, that is. This is because there’s considerable investment going into some (parts) of us from both government, commercial and third sector organisations who are banking on us being shaped as they wish through all the big data we will be sucking up and spitting out. We relish the attention some of us are being given and the ‘smart’ digitization that comes with that. We like to be tickled with your digital dramas. They are entertaining and possibly also a little bit addictive. But that doesn’t mean it is good for us. Too much of a good thing doesn’t usually end well.

We are bemused by why the more big data you have to model us, the more you treat us as though we are but a series of links, circuits and networks. You don’t seem to do this for other species. For example, flocks of birds still remain about ‘birds’ and schools of fish are still fundamentally about ‘fish’, even if the methodological approaches are sometimes shared. Yet when it comes to the way you self-organise into cities and urban regions, you seem to lose sight of the fact that we are fundamentally about humans – or rather human ‘being and becomings’ (Uprichard 2007), being in the present and always changing? Why does big data seem to strip away the humanity that is fundamentally at our core? There seems to be a blind spot where you fail to see your own human reflections in amidst the new digital tools and mathematical models. Big data might well be a growing part of us, but we are still mainly social because of you. Cities – smart or not – are only here because that is how you humans self-organise yourselves through time and space. Surely big data should be helping you to form the emergent structures you desire?

2. We need more ‘big’ methodologies, not more big data. You may not like this point because many of you tend to be set in your particular epistemological ways. We find it very difficult to speak to everyone together at the same time in a way that won’t also vex many of you; we do not mean to offend or upset, but merely set up a dialogue about this particular point of (mostly methodological) contention. Put simply, big data has the potential to show something new or interesting. However, we are increasingly worried that you are missing the bigger picture of how we work – and the way we work is through you, not through data per se. So whilst at one level we accept that big data can show certain circulatory and transport systems better, we are unclear how that helps us to choose particular kinds of trajectories holistically. We are not convinced that you appreciate how much coaxing we may need to change in particular ways and that your current methodological approaches to big data may be alienating us further rather than bringing us together.

Please don’t get us wrong, we generally value all the methods you currently use. And goodness, you have developed so many! Cellular automata, agent based modelling, mathematical and computation modelling, networks, participatory GIS and mobile methods, right through to more traditional approaches such as surveys, interviews, ethnography, documentary analysis, architectural planning and design and so on. It’s impossible to list them all; we don’t even know all of them as they don’t always get used on us all everywhere or in the same ways. Indeed this is partly the problem; there are many ways to know us. Big data methods are but one family of approaches.

But whichever methods are used, in whatever mix they are deployed, they can arguably only end up producing particular political ‘cuts’, as Karen Barad would call them. For us, these ‘cuts’ don’t always stitch together meaningfully even if they can be seen to be ‘force fitted’ sometimes. We rarely get asked what we think about particular methodological cuts let alone how they get sewn together. What worries us is that you seem to increasingly want to join these cuts up neatly and yet some of our best parts are old and messy, knotty and awkward, grubby and yucky. Your methodological cuts don’t have to fit together well to be right, and nor are they right simply because they do. What we prefer are empirical cuts that allow us to show off how and why we are growing and changing in particular ways and not in others. We can give you all the data you like, but unless the data layers are sewn together in a way that helps us move and stretch around, we don’t see how we can ever come close to showing you how fast or slow we can move and morph through time and space – or, for that matter, why and how we might not be able to change in the ways you want us to.

One of our worries about big data and the analytical orientations they nurture, is that they can make us seem as though we are all the same. This is not to say that we don’t have some features that make us look alike. We can certainly be lumped together into pixel bytes and storage stacks. You capture snap shots of us very well. The same kinds of machines, variables and methods are used to process much of the big data visually, so it is understandable that the spatial images tend to be of similar textures too. Satellite images, various chloropleth maps, and animated data points are part of the image library that is emerging from big urban data.

These big data spatial snap shots are akin to ‘urban selfies’, which each capture a public, playful part of us. Urban selfies may be fun and fine, but en masse, they do two main things. The first is that they render us generic. We begin to resemble each other more than we differ. ‘Urban selfies’ make it difficult to tell us apart because we are posed, dis/coloured and animated in almost identical ways. Yet it is our very local and contingent differences that matter in terms of explaining how we are differently constituted and changing.

Another concern is that, if left to the uncritical eye, these images can seem to capture something more real and tangible than ourselves. In other words, the urban images about us come to be/come us. The ontological and epistemological dimensions of the smart cities are subsumed within the methodological possibilities of big data. One way of recapturing the city through the data driven images may be return to what may be particular about urban description more generally. After all, inscribed in every empirical description of the urban is a history and context of why those data have been collected and not others, a set of actors – human or otherwise; information about the intentionality and processes, as well as information about the unintentional processes, are encoded within them. As Lefebvre (1974/2003:74) put it, quoting Heidegger, whilst a rose does not know that it is a rose, ‘a city does not present itself in the same way as a flower, ignorant of its own beauty’ since it has ‘been ‘composed’ by people, by well-defined groups.’

So, by implication, what is also shown – by virtue of omission more than through the analytics themselves – is how unequal data access and production are, where similar kinds of urban socio-economic polarisation may or not occur, and how powerful some agencies are making themselves more in/visible. Big data doesn’t necessarily help us to make decisions about what to do about that and nor do the images big urban data produces. Therefore, how exactly these urban selfies help to lessen the gap of inequality among us remains unclear to us, even though we can imagine how they might be mined to do something different to what they are doing already.

That said, we understand the convenience of blurring our differences and thinking of us as all the same. That way, the giant global companies that already have such a stronghold on all of us can simply roll out ‘one size fits all’ initiatives. But, like you, we are each unique. Difference really matters when it comes to causality, which is complex and equifinal. Because we can reach similar states via so many routes, understanding the differences in why we may or may not change the way you want us to really matters. Moreover, it is in the different pathways to the same the outcome that you may also see how reluctant or willing we may be to change in line with your plans.

Despite our spatiality, which is what the big data images emphasise, we are as much about time as we are about space. It is what makes us so mischievous and so brilliant at continually adapting. Our different trajectories and tempos are what we would like you to see, as well as the parts of us that may not have the propensity to change qualitatively. That is why we would prefer big methodological approaches that render more visible our multi-level and multi-dimensional trajectories in terms of the differences and similarities. Such approaches can help us tell you more about the temporal dynamics of our very being and becoming. Big data may be able to do that, but the level of granularity that is useful for commercial entities may be a very different one to the one that will help you at the level of policy. Questions about who benefits from big data descriptions may need to be explored together. Likewise, it may be useful to re/visit together an examination of the kinds of big data driven descriptions that may maximise and/or hinder policy, planning and practice.

3. We love and hate being called ‘smart’. Being called ‘smart’ makes us feel wired up, as if we are becoming a bit more human. Perhaps we can even become more like you; it is flattering to think you think we can be more like you and more brain-like. Being ‘smart’ is certainly a handy title since it helps some of the researchers and planners who have worked hard, and the businesses that have pumped money into us, to make us ‘smart’ to also feel a bit clever and ingenious and it’s good to encourage them to keep nurturing us in various ways. The problem is, calling us ‘smart’ implies that all our neighbouring urban and rural regions may be ‘stupid’. But many of these ‘minor’ hubs are some of our silent powerhouses and local and regional drivers. They are what enable us to be and become as ‘smart’ as we are. We are completely inter-dependent. But they are beginning to withdraw and disengage with some of us, as are you with them – this will lead to nowhere good, especially if you want us to be able to feed and look after so many people who will end up living within us. To be ‘smart’ with and about big urban data seems to imply that the areas that are not as ‘data intensive’ are ‘lesser’ in all kinds of ways – less important, less worthy of investment, less demanding and so on. But this is wrong. Being ‘smart’ is beginning to make us all look a bit stupid and that’s rather embarrassing.”

Beauregard, R. (1995) ‘If Only the City could Speak: The Politics of Representation’. In H.Liggett and D. Perry (Eds.). Spatial Practices. London: Sage. pp. 59-80.
Harvey, D. (2000) Spaces of Hope. Edinburgh: Edinburgh University Press.
Kauffman, S. (2000) Investigations. Oxford: Oxford University Press.
Lefèbvre, H. (1974/2003) The Production of Space. Oxford: Blackwell.
Soja, E. (1980) ‘The Socio-Spatial Dialectic’. Annals of the Association of American Geographers. 70: 207-225.
Tilley, C. (1994) A Phenomenology of Landscape. Oxford: Berg.
Uprichard, E. (2008) ‘Children as Being and Becomings: Children, Childhood and Temporality’, Children and Society, 22(4): 303–313.


Emma Uprichard is Associate Professor and Deputy Director at the Centre for Interdisciplinary Methodologies and Co-Director of the Warwick Q-Step Centre, at the University of Warwick. She has recently completed the ESRC project ‘Food Matters’, which explored food hates and avoidances through the life course and is currently on the ESRC Seminar Series on ‘Complexity and Methods in the Social Sciences: An interdisciplinary approach’.