I came to grad school to have good conversations, particularly about what it means to design data visualizations. But I discovered that data visualization has the possibility to help us in even more critical conversations.
It started for me with the people in my everyday life. In the crushing stress of a demanding grad school program, we all had a grasp on these intangibles in the air: the sense that something was not right; whether we felt alone, stressed out of our minds, the idea that we had unhealthy coping patterns. This is the feeling communities must address in a world filled with climate change, systematic oppression, and pandemics.
In communities, experiencing crisis or not, conversation is the key to solidarity and belonging. Scholar Peter Block writes that the small group conversation is the site of community transformation (Block, 2008).
Thus, we must design for these transformative, community conversations. So I began holding weekly conversations with my peers.
How could I, using my data visualization background, support conversations such as these? Data visualization, as a medium, is uniquely suited for supporting these kinds of convos. Paraphrasing Alberto Cairo, prominent data viz scholar, data visualizations create good conversations by showing what we know visually. They make what we know, which can sometimes be intangible, into visible and discussable patterns.
Community conversations expose a crisis in how we design data visualizations.
In a former life, I might have addressed the visualization needs as follows.
I, the data visualization designer (perhaps with other subject matter experts), would design a data collection survey, assemble my community, and collect the data by asking them to fill out the survey.
I might provide one dimension to ask how folks were feeling. They might think “what is the correct interpretation?”
Data collection here is a cognitive activity. The data collection process lends itself to thinking through each of these steps logically, not necessarily tapping into the feelings in their body — the tightness of the chest, the nausea in their stomach — perhaps trying to ascertain the one true interpretation of my survey. Their felt experience is ignored, and multiple interpretations of the survey are lost in exchange for consistency. In sum, this data collection process feels extractive.
And so after all of the surveys are collected, I, the data visualization designer crank the numbers and produce this dashboard of “our community.”
This is the kind of conversation we might have: First, I might assess the patterns in the data visualization.The conversation might continue. The question in this conversation might be: “how can we get more people to be happy in the studio?”
That focus comes from these typical characteristics of data visualization itself: for example, these circles and rectangles abstract away the people (and their stressed bodies) they represent. Even further, the people in the community are very much isolated in the visual despite being very much entangled in the social texture of the community at large. Finally, the “truthy” aesthetic has no member questioning whether there are any alternative views of the same community. We may find ourselves taking this truthy aesthetic for granted.
Notice then, the ways in which this typical data visualization approach changed how this community talks to one another. This visualization leaves us in the posture of analyzing these nodes and patterns, trying to control them. Meanwhile, it keeps us emotionally at a distance, leaving the tightness in our chest undiscussed and festering.
This speaks to concerns scholars have about data visualization. In particular, there is an emerging sense among scholars of data visualization that our practices create rigid, harmful understandings of the world. Digital humanities scholars have addressed this (for example, Johanna Drucker’s work on capta). Further, a wave of artists, practitioners, and scholars have been building on this thread, dissecting the implications of practicing data science and visualization in a disembodied way: Catherine D’Ignazio and Lauren Klien’s Data Feminism project, Jenny Odell’s writing on data and Designing for the In-Between, Safiya Umoja Noble’s work on Algorithms of Oppression, and Mimi Onouha’s work on data collection and ethics among others.
Our practices of data visualization can have potently negative consequences. D’Ignazio and Klein argue in their book, Data Feminism, that current practices of data visualization obscure bodies, leading to the silencing, extraction, monetization, or invisibility of people (or bodies) that the data represent.
As put by Jenny Odell in a 2018 talk at KIKK for “Designing for the In-Between”:
“There’s so much that stands to be rendered invisible by any system of knowledge or identity simply because it doesn’t fall squarely into one bucket or the other.”
This is a jarring realization as someone who wants to help make a better world. Not only am I not cultivating a helpful conversation, my data visualization efforts might even be damaging. WIth this sample data visualization’s conversation, we are left relating to one another largely through analysis, logic. I realized that my data visualization education and experience leaves me without theories and practices that help me make data visualizations in a helpful way.
The theories and practices missing in our approach to designing data visualizations.
Data visualization is ontological
I realized that this relationship between data visualization and our daily life that makes data viz ontological. That is, data viz impacts how we talk about the world, how we understand the world, and thus, how we relate to it.
I did some soul searching about my data viz practice; something needed to change. With this ontological approach in mind, then, there must be a different way of collecting and visualizing data, beyond this relationship of extraction, analysis, and control. That’s the space this thesis project has been carving out.
Bring back the bodies
What is missing in our typical approach to data visualization, then?? In the words of Catherine D’Ignazio, We need to make visible those who are creating the data, who are represented by the data: we need to bring back the bodies.
Exploring how our theories and practices of data viz change when we bring back the bodies.
My project asks: how do our theories and practices of data viz change when we bring back the bodies? I began designing data visualization experiments for these same critical community conversations to find out. I broke my explorations down into the fundamental questions designers implicitly (and explicitly) answer when they collect data and create data visualizations.
This is what my thesis project explorations looked like. Imagine you are a member of my cohort, and we have gathered for an open and honest conversation of how we are doing this week.
Data collection becomes about designing for experiences for reflection and noticing the self
[diagram of data collection process]
I openly disclose that we’ll be training a neural network on your phone and how the data will be stored and used, using your hand gesture to capture how you’re doing. I then introduce the way we are collecting data, saying something like: Today we’re going to reflect on our community experience through the lens of the tree’s root. Like us, roots spread and grow when the body is strong; they don’t grow and spread out as readily when the body is weak.
Imagine this tree represents you. Hold up your hand to the camera. Imagine you’re a tree root growing — reaching out into the soil for nutrients. Reaching, grasping, opening up Imagine you’re a tree root retreating, contracting. Slowly slowly close your hand, millimeter by millimeter. Contracting, closing, turning inwards.
Take a moment to find stillness, meditating on this motion of opening and closing. Find the hand position somewhere here that matches how your body feels. In my prototype, you’d be able to submit a hand position anywhere from closed to spread.
Here, data collection becomes about reflection. the body and what it feels becomes the center of attention — that tightness in the chest, the nausea in the stomach. The scale of meaning is tied to your unique body, so multiple interpretations of the gesture are built into the framing of the data.
Designing data visualization becomes about designing the experience of beholding the world
And when we all submit, each of our trees grows together, in a forest metaphorically representing our community.
And then I might tell the group of people this about the design of the data viz: Like humans, trees can only thrive and grow old with the support of those around us. Together, they create the conditions within which they can thrive. Inspired by Wollheben, I’ve created this data viz, imagining that we can help each other thrive like the old growth forests of Pennsylvania do [1].
I say this to my community and we look at our visualization. Each of those hand gestures submitted by phone is sent to a forest that grows roots.
[diagram of visualization]
The conversation becomes about working together
The exact meaning or value is intentionally ambiguous, so the conversation starts there. Another member says “looks like many of us aren’t thriving. What’s up?”
The conversation might continue, with folks looking at different interpretations of how they’re thriving or not. And sort of the main question in this conversation, based on my visualization, might be “how can we work together to thrive together?”
Designing for this conversation has revealed to me new theories and practices of data visualization
While not all conversations will be about the feeling of people in their communities, my project designing for these conversations reveals many important ways we need to bring back the bodies as we design with data.
It starts with how we define data. I take a feminist perspective, also inspired by object oriented ontology, in my definition of data. With this take, data is the phenomena observed by bodies. This means that data does not exist in a vacuum — it’s observation is bound up with the observer. The felt experience of the body — emotion, the tightness in the chest, etc — is valid — and crucial as data.
Because data is phenomena observed by bodies, we need to attend to how the body understands the world — through physical motion over time.
Finally, because each body, each person, knows about the world in different ways, we have to design for multiple interpretations. Ambiguity helps with this.
I’ve given examples of two reasonable visualizations of the same community, and two radically different conversations. These are some ways I’ve uncovered for bringing the body back in data visualization.
[aggregated list of practices / principles outlined (lists with I, II , III, etc.)]
They are non-exhaustive, but they already show the potential for turning data visualization away from a tool of control towards one of transformation. At the end of this thesis I feel as though I can contribute to critical conversations as a data visualization designer.
Conclusion: an urgent call to data visualization designers
When we look at the kinds of failings we have in the conversations about our data visualization, particularly during this pandemic, the need to address how data visualization impacts how we relate to the world is critical. To develop new practices based on these ways of bringing the bodies back to data visualization.
What can a good data visualization do? Or, what can be accomplished with ontological data visualization?
Good data visualizations enable us have transformative conversations by holding space for multiple ways ways of knowing about the world (ambiguity), giving us a chance to viscerally understand, or have a feel of, the world and giving us a visual record of our intangible patterns How we might do this:
- Reflect on your perspective as a designer. How do you relate to your data visualization work?
- Humility in your perspective and the work you produce
Lean into your responsibility and commitment to those you are visualizing and for whom.
