Reframing our relationship with data and living it
The framework of this essay was originally presented as a short talk at the Visualising Data London meet-up on 24 January2019.
It’s often said that a picture paints a thousand words.
But what if that image needs to communicate vital and complex information? Or tell an extremely important story? If it does, what happens if it is too abstract? And what if, despite trying its best to convey it, it is void of emotion?
If an image fails to build an empathetic or emotional bridge with its intended viewer, does it have any chance of fulfilling its objective? Will it communicate the desired message successfully? Will it inspire people to act?
These questions strike at the heart of what we practitioners do within the data visualisation space, and they call into question our field’s core objective. If we answer them collectively, I believe it will move us forward.
I’m not suggesting that we throw out the rule book, I’m encouraging us to actively question it. We need rules and guidelines to help us strive for transparency, accuracy and “factfulness” as defined and given to us by Hans Rosling. But, I’m worried that the formal attributes of traditional data visualisation methods are hindering us from connecting and communicating with people effectively on an emotional level.
When the intent is to present data to inspire action and facilitate change, we should reconsider its delivery. Of course, this doesn’t apply to all data communications and data types, but it’s an interesting problem to dig into, and is very much a core part of what we do at DATA4CHANGE.
These are some of the items from my reading shelf that helped me formulate and begin to explore these questions.
Each of these works offers us a unique perspective on how we might start to reframe our relationship with data.
What they do profoundly well is challenge established conventions about how we look at, use and approach data as an entity, subject matter and content type. Each one touches data through a unique perspective and on various levels: from data collection, to methodology, to metrics and indicators, to the translation of it, right down to how we formulate insights from data and translate this into consumable or information.
Connecting the dots
Where the ideas above merge into a single consolidated focal point is in the brilliantly complicated book Being Ecological (2018), by Timothy Morton.
It is, of course, a book about ecology, but it’s about so much more than that. It interweaves a relatively new theoretical philosophy of everything called Object-Oriented Ontology, or OOO (“Triple O”), with ecology, to help us better understand our relationship(s) with the world. It is super complicated, but on page 15 sits the sentence that launched this inquisition of mine and ultimately led to the questioning and interrogation of these words: data visualisation.
Here it is:
As I understand it, Morton says we must “start to live the data” before we can begin to comprehend what the data is trying to tell us. This is further compounded by the way the data is typically delivered, which is currently hindering our ability to understand the complexity and urgency of it.
Seemingly simple. Live the data…but how?!
The problem isn’t that people haven’t been given enough information, stats, facts and charts, but that they haven’t been delivered (designed) in such a way that inspires an appropriate reaction and connection. We need this connection in order to be inspired to take action on the information we receive.
Morton’s writing is about climate change data, but if this is true of an issue that is inherently human-centered like climate change, the application of this principle has a much broader scope.
What Morton adds to the landscape of this discourse is how we go about expressing (visual or otherwise) the information we extract from data.
Data expressions
This is a term that you’ll find through the rest of this essay, and is the key concept I’d like to explore. I’m not sure where I’ve come across this term as defined in this particular context, and it seems too obvious to have not been defined before. I’m aware of its existence in the data science and programming spaces, and will happily update this essay with references if pointed in the direction of work addressing this elsewhere.
Inclusivity is accessibility
At the moment, I’m finding those two words data visualisation to be extremely limiting and exclusive.
Data visualisations often rely on one single dimension of sensory input: sight.
But a data expression could deploy any sort of sensory expression: sight, smell, taste, touch, sound. Singularly relied upon, each of these senses of course have their own limitations and exclusions. It isn’t until we begin to mix and match them that we start to find more inclusive expressions of data.
We often see the use of sight and sound combined to create the classic “data visualisation interactive”. I hope that data-driven expression and data-driven content could become easier to understand and access by using more sensory inputs. If it is understood and accessible, it is much easier to inspire action.
So, let’s look at what could be considered a typical pipeline for distilling raw data into a visual data expression.
The Pipeline: Data, Information, Knowledge, Wisdom
By dissecting the Data ->Information ->Knowledge ->Wisdom pipeline, we might be able to identify key moments where we can layer in other sensory inputs.
Let’s start with data.
We often wind up with a raw data set or a set of summary data. We begin by getting our heads wrapped around its content and context. We ask, what is it telling us? What’s the nature of it? How might we start to explore it to gain further insight?
What we don’t often do, is spend much time thinking about where the data came from.
Who collected it? Why did they collect it? Was their motivation to ultimately prove or disprove something with the data they collected? Who designed the research methodology? And what are their biases?
We cannot divorce a dataset from these questions. Data, as we know it, is a mere abstraction of a series of interactions that have been captured by tools that humans have built and these interactions are more often than not driven by human relationships.
Sound familiar? I hope so! Because this concept is one of the key pillars in data humanism. It’s a perspective that people like Hans Rosling, Giorgia Lupi, Catherine D’Ignazio and Lauren Klein have been championing for quite some time.
When we are able to tap into a dataset’s core context and subject matter, we will be able to unlock the narratives behind the numbers. We will explore these preliminary insights and expand upon them as we progress through the pipeline, eventually using them in a data expression that has a greater potential to encourage human connection and build an emotive bridge.
Information
Next, we move on to our initial exploration of the data. We test it, question it, slice and dice it, examine it. We begin to turn our data into information, which we can then begin to explore further. Once we’ve done this, we can really begin to better understand the nature of the data and listen to it. If we listen, look and observe closely enough, we can begin to translate these key insights into other forms of sensory input and go beyond ‘the visual’.
Knowledge
We add yet another layer of subjectivity and opportunity when we move from information to knowledge. To gain knowledge, we link information to other contextually relevant pieces of information. We create relationships and form perspectives. Then, we start translating these core perspectives and other contextual bits of information into sensory outputs.
Wisdom
When I work on a project, I’m hoping that after having spent all those hours crafting an expression of the original data, that the expression will mean something to someone. I want them to access it, to understand it, and to put that ‘wisdom’ to use. My lofty aspiration is to create something that is a catalyst for someone to transform their knowledge into actionable wisdom, but experience shows that this is the most difficult thing to achieve.
For your consideration
Here are two relevant examples that speak directly to what has been discussed above.
Waterlicht
Dutch artist Daan Roosegaarde’s Waterlicht (Water Light) has been installed in multiple locations since its inception. It floods a site specific location with waves of light that reflect rising sea levels. The Water Light project demonstrates how the Netherlands could potentially flood without the protection of its water infrastructure.
Ice watch
The installation Ice Watch from Olafur Eliasson empathetically confronts its viewer with polarizing effect. The viewer is presented with what are seemingly harmless hunks of ice but in reality, the “Twelve large blocks of ice cast off from the Greenland ice sheet are harvested from a fjord outside Nuuk and presented in a clock formation in a prominent public place. The work raises awareness of climate change by providing a direct and tangible experience of the reality of melting arctic ice.”
A small child might not be able to read a chart per se, but they are capable of feeling, seeing and even tasting an expression of data, which they then might ask their parents about, which would foster a deeper conversation, dialogue and human connection around these important and pressing issues.
So, how can we get people to act?
This is a question we are actively exploring and investigating at DATA4CHANGE in our daily work and at our events. We have a long way to go, a lot more people to talk to, and a pile more books to read and perspectives to consider, but we have to start somewhere. We’re hoping that by opening up and contributing to these discussions, that we can both shape, and be shaped by, radical thought that changes our practice for the better.
