I was invited to give the opening address at the 2017 graduate exhibition of Density Design, Politecnico di Milano, one of the world’s leading research labs focusing on the visualisation of complex phenomenon.
This year’s graduates produced a very interesting set of projects based on both official data and web data – on topics from climate change adaptation to climate finance, cultural heritage to migration.
Data Visualisation, Data Experiences and Data Worlds
First of all thank you very much to Paolo and Michele for the kind invitation to give this address at the Density Design Lab’s 2017 graduate exhibition in Milan, which I consider a great privilege given the pioneering work that your group has produced over the course of these past few years.
Today I will aim to provide a bit of reflection about the significance of data visualisation and its role in shaping the future of the data societies that we live in. And of course, an appreciation of the politics of data is surely both even more important and even more challenging in the so-called “post-truth” era that we are said to now live in.
In a manifesto for the Guardian newspaper in 1921, one of its editors C. P. Scott famously wrote that “comment is free, but facts are sacred” – a slogan which has become popularised through the Guardian’s Datablog.
“Facts are sacred”: How hopeful and optimistic this phrase sounds after the past year, a year in which we were told that citizens had “had enough of experts”, a year when facts – facts about everything from climate change to migration, economic trends to political events – have been challenged, contested, politicised, and are no longer taken for granted.
As well as established institutions of fact-making being challenged, we have also seen the rise of other cultures and ways of viewing information from the “alternative facts” of Trump Administration Counselor Kellyanne Conway, to the production and circulation of so-called “fake news” in countries around the world.
In his final speech before leaving office, former Vice President Joe Biden, spoke of the “dangerous proposition that facts no longer matter”, suggesting that this posed a threat to “the liberal international order”, which could indeed collapse, leading to “a world where the strong impose their will through military might”.
What kind of role might data and data visualisations play in this post-truth, post-fact world? How might these developments lead us to think differently about data and data visualisations – and their role in collective life?
As you may know, the word “data” comes from the Latin “datum” which means “(thing) given”. However as I’m sure that you’ve learned from your various projects, datasets can often be messy, complicated, contingent, and not at all straightforwardly “given”.
There is often quite a lot of work – “data work” – that is required to assemble, clean, align, analyse, interpret and transform datasets, not to mention the social and political work that goes into making and building consensus and agreement around them.
While Edward Tufte says we should “let the data speak for itself”, as I’m sure you’ve found in your projects it is not always exactly clear what the data would like say, nor the best way to say this.
While we should indeed always spend time “listening” to our data, the process of figuring out how to find meaning, narrative and experience in this material often unfolds as a process of negotiation between different elements – including people and software, visual forms and code.
As I’m sure you’ve found, this process of negotiation can be just as much an art and a craft as a science. And rather than taking the data that this craft depends on as “given”, it is important to consider the things and processes upon which data itself depends.
The historian of science Theodore Porter characterises the creation of quantitative data about the world in terms of an aspiration for what he calls “thin description” – which involves various kinds of work to make different aspects of the world comparable across space and time, and to render different aspects of the world as quantities – whether through measuring, counting, aggregating or evaluating.
These “thin descriptions” of course carry enormous advantages. Attention can be drawn to very specific aspects of diverse and complex phenomena – such that our attention can be restricted to just a single number, or a single point letting other aspects of the phenomenon recede into the background.
It is precisely the thinness of these forms of quantification and datafication that makes it possible to see and experience patterns and dynamics that might otherwise be difficult to detect – such as in the paradigmatic case of Jon Snow’s cholera map.
But making data is not simply about representing different aspects of the world – but articulating and performing the world in different ways. As Ian Hacking argues, data contributes not only to measuring populations, but also to making them – as with the case of the classification of diseases such as “multiple personality”, which not only represented a new form of illness, but which also created a new social group who could be said to be suffering from it. In this sense thin descriptions may be considered productive.
For example, when we create indicators or rankings about ideas and values such as “openness” “liveability”, “innovation”, “competitiveness” and “sustainability”, we don’t just pick out pre-existing aspects of the world, but create the conditions of possibility for seeing, experiencing, understanding, responding to and interacting with the world in certain ways. Data not only represents different bits of the world, but contributes to making different kinds of worlds possible.
And data visualisations have a very important role in the mediation of these “data worlds”. Data visualisations may be considered as part of the front line of datafication. They make possible certain kinds of visual experience of data – not only by rendering different patterns and relations observable, but by giving us a kind of visual language with which to think with, through and about data.
It was a pleasure to peruse through all of your different projects and see all of the different visual forms which were used in order to create experiences and stories with data – including combining maps, alluvial diagrams, charts, grids, networks, images and of course bubbles – or “pallozzi” as I hear that they are called in Milan.
There are two kinds of “worlding” or “world-making” which can be noted in relation to this year’s exhibition. Firstly there are the world-making capacities of your projects issuing from different styles of visual reasoning, visual analysis and visual experience.
Just as the invention of the timeline or the network diagram facilitated styles of reasoning and experience with data which had not previously been possible, in a similar vein this exhibition shows that Density Designers are defining new forms of storytelling, new ways of bringing together data, image, text and code into web-based interactive – and new genres of “data experience” – through their work.
One thing that I particularly enjoyed in many of this year’s projects is the question and answer based work-flow diagrams that show your reasoning and the “data work” behind the visualisations, including the data, methods and approaches that were used – and how the data was generated.
These are very useful as a way to show what is involved in the data visualisation process – and to cultivate a broader critical literacy around what is involved in making the “data worlds” that you explore in your work.
I’ve been collaborating with several researchers on the use of the network as a visual form which is used to work with, reason with and display data in journalism and social research. One thing that we’ve learned is that networks are used in quite different ways in different settings.
For example, while researchers often use networks as exploratory devices to discover patterns of relations, journalists often use them as explanatory devices in order to clearly demonstrate hidden ties or unexpected connections for their audiences. These different modes come with different methods, ways of working and ways of seeing with networks.
While we may tempted to see vision as something that is direct, spontaneous and natural, it is important to remember the different kinds of cultures and traditions that seeing depends on.
The late John Berger wrote about this in his masterpiece Ways of Seeing, and I’d argue that many of his arguments and insights about visual phenomenon such as fine art and advertising also apply to the ways of seeing associated with data visualisations.
Another perspective on visual world-making is available in Scott McCloud’s excellent Understanding Comics which explores the worlds of experience and narration which are made possible through the different visual forms and conventions of comic strips.
As he points out, the ways that space and time are articulated through the boxes and speech bubbles that are so familiar to us is extremely interesting and not always intuitive.
In the future, I’m sure that researchers will look back to the archives and the legacy of Density Design as an important source to understand the culture, craft and aesthetics of making worlds and experiences with data in this period.
The second form of world-making which I’ll mention: I thought it was striking that so many of the projects were transnational in their scope and focus – from climate change to migration, culture to terror.
There is, of course, a long tradition of exploring collective life at a planetary scale in data visualisations – from the internationalist experiments of Otto and Marie Neurath and their Isotype Institute in the first half of the 20th century, to the late Hans Rosling’s work looking at global trends with the Gapminder Foundation’s Trendalyzer tool.
Through so many of the projects in this year’s exhibition, one becomes aware of the relations, dependencies, differences and commonalities between and across countries – making visible the planetary scale patterns which we are not normally able to see.
Just as the “Earthrise” image taken from Apollo 8 occupied such an prominent role in the popular and political imagination in the 1960s, so data visualisations have taken on a crucial role in relation to how we understand collective life in the 21st century.
Here I’d emphasise the role of data visualisation in relation to other projects and practices of world-making in this second, much more literal sense of planetary scale coordination and integration – from map-making practices, to global statistics; global data flows to international institutions focused on everything from cultural heritage to migration to energy.
Here data visualisations help not only to represent and communicate information, but also to undertake rhetorical, cultural, narrative, social and political work in relation to the collective task that sociologist Bruno Latour describes as “composing a common world”.
As he puts it – we are in a plane from the “land” of previous ages, en route to an imaginary “globe” which we cannot reach due to the limitations of our planet. So where can we go? In the midst of backlashes to globalising projects in many countries around the world, how can we redefine a new path for collective life in this comparatively narrow band of inhabitable space that geologists call the “critical zone”?
Finding ways to explore our interdependencies, to render connections and facilitate shared understanding and bonds of solidarity between people in different places has surely becoming a defining task of our time particularly in the face of a range of urgent transnational challenges (such as the potentially catastrophic impacts of climate change), as well as in response to the current political trajectory of the US and many other countries.
This year’s projects demonstrate a range of ways to “take measure” of the world through data – which brings me to my final point. It is crucial that we don’t lose sight of the question of what is measured, how and to what end. Just as it takes a lot of work to make data visualisations, so too it requires work to create data. What data is collected in turn shapes the different kinds of data visualisations and data experiences, analysis and story-telling which are possible.
We are in the midst of some major changes in the global production of data – not least because of new forms of “born digital” data. It is very good to see that with this year’s projects you have looked at both “official data” and “web data” – in order to show what is possible with these different sources. Digital technologies make it possible for people to not only use data, but also to participate in the making and shaping of data.
So my final reflection to you all is to think about how data visualisations facilitate not only public communication and public understanding, but also more substantive forms of public engagement, involvement and participation.
Here we may also benefit from debates between John Dewey and Walter Lippmann in the 20th century – who said that there is essentially no such thing as “the public” (which is an abstraction), but rather many different publics.
What becomes essential with this shift is to understand how these publics are formed, and the different mechanisms through which they are able to organise and mobilise around issues. As the sociologist Noortje Marres puts it “no issue, no public”.
Drawing on this kind of analysis, more recently people have discussed “data publics” – and so the thought I will leave you with is to consider the role that data visualisations can play in assembling and supporting the capacities of these data publics. Different kinds of media facilitate the gathering, mobilisation and movement of different publics in different ways.
Given that Density Design graduates have gone on to do such important work in advancing the state of the data visualisation field at organisations and institutions all over the world – I’ll take the opportunity to plant a seed for you, for whatever you do next, and wherever you go, in the form of a final question.
What role might data visualisations play in developing capacities not only to use and understand data, but also the capacities to actively participate in shaping the future of the data society? In other words, how can the craft of data visualisation develop so as to include, engage and involve people in processes of making data and understanding how data is made – so as to support them in becoming data citizens rather than just data subjects?
Congratulations to you all on an outstanding set of projects and I wish all of you the very best of luck with your future endeavours.