What happens when we dance our data?

by Victoria Palacin, Laura Perovich, Rosalie Norris, Rahul Bhargava 

More and more decisions each day are made using data. Data scientists and policy makers are usually the only ones with the expertise to engage with data. If we want to empower others to find and tell stories with data, we need to create new ways to engage with information. Over the past decade, we have been exploring community engagement in a variety of creative ways, including Laura’s interactive data experiences, Rosalie’s creative invitations to dance, Rahul’s large-scale community data murals, and Victoria’s community technology creations. But there’s still more to be done if we really want to make data an asset people can use to make their communities better. 

In this spirit, 6 months ago we decided to see what would happen if we asked people to dance their data. How could we prompt regular folks to perform the stories they find in their data? Would that change how they relate to it? Could it help empower them to engage with data they otherwise would feel intimidated by? This short, slightly-academic blog post summarizes some of our first efforts and the new directions we found through the beginning of our community focused design process.

cc Ahmad Odeh

but first, some background

An ever-increasing amount of data about various aspects of the world around us is recorded. Many believe that the path to ideal decision making lies in this data and associated analytic methodologies. However data is far from neutral; it is highly contextual and carries many limitations. In order to interrogate data and understand its perspectives and biases we need easily accessible data tools and methods for collective sense-making. As we move towards more automated societies where data is a de-facto part of our infrastructure, isn’t it fair that we, the people, are allowed to make sense of our own data? 

We are living in what many regard as a “post-truth” era, with declining civic participation, structural discriminatory discourses on the rise, and conspiracy theories challenging scientific knowledge. Creating data literacy skills is far from being the magic solution to these issues.  But, if we are to start somewhere, shouldn’t we think of enhancing the capacity of people to understand, question, and relate with data? We should understand how and what data is being collected around and about us, why is it meaningful to us, and what it could be used for. This process of collective understanding can foster trust, empathy and creativity.  Data can become a means for insight, evaluation and reflection. For this to happen data must be brought into the public sphere.

Here was our modest idea to help – how about we let people make sense of the data in their communities? Of course, this is easier said than done. Currently, the tools and methods to facilitate data sense-making without all the technical jargon are still limited. 

thinking about how we interact with data

We learn to interact with the world around us through touch and movement from the moment we’re born. Yet most work around data is centered on screens and spreadsheets that remove context and limit our ability to engage with the information in an embodied way, relate it to the world around us, and create positive change. 

Some research to tackle this has been growing in the fields of data literacy, physicalized data, and somaestetic design. Rahul’s Data Culture Project (created with Catherine D’Ignazio) has developed a series of fun and creative activities and tools to help organizations create a data culture. The data physicalization field explores how to express data through physical forms instead of spreadsheets or digital tools, their examples range from pre-colonial ways to store data such as Incan quipus to modern exciting and creative artifacts like tattoos, cuisines and rollercoasters (numerous examples catalogued). Somaestetic design focuses on designing experiences that augment our understanding and engagement with oneself. This is done by engaging people to shift their focus from external sensory interactions to deeper body experiences like sensations and movements (Höök et al. 2015). Scientists, educators and policymakers have also used dance for communication, for enhancing scientific dissemination (e.g. the Dance Your PhD competition), improving learning (e.g. dancing statistics, dancing classrooms), nurturing empathy and shared identities (e.g. dance and dialogue, police squad dancing challenges) and guiding political campaigning. There’s also the Theater of the Oppressed’s important historical work using theater to understand challenging situations, create collective learning, and move towards positive social change. 

With all this inspiration in mind, we decided to work on a project we call “Data Moves”. We are seeking to change the way we present and interact with datasets, using dancing and acting as vehicles to tell data stories. As such, Data Moves builds on Laura’s idea of a data experience, which “takes data off the screen and puts it into the physical world” (Perovich L., 2015). 

Why dancing and acting? Because dancing and acting have a long history of being used as effective means for conflict resolution and empathy building. The human body can become the embodiment of stories grounded in data. The goal is not necessarily to produce a well crafted performance, but to collectively explore and dig into a data set; using our own bodies to tell the stories we find in it and the way we feel about it. In this sense, dancing and acting are methods that can bring people and data closer and open room for creativity, fun, and reflection. A more emotionally engaged approach to data may also open our minds to new discoveries and understandings.

We have been seeking to create experiences that don’t distract us from our embodied selves and our real world contexts but instead deepen our understanding and engagement with ourselves and the data we are exploring. Critically and reflectively asking how we feel about the data can provide important insights into the social phenomena and policies that may be concealed in the dataset.  

dancing our data: initial explorations

We have run two Data Moves workshops. Both workshops were run in partnership between academic and a dance educator/community organizer. Collectively the organizing team brings together knowledge in collaborative design, community engagement, data for all, civic tech, dance, data visualization, and theater. The outline of the workshops was roughly the same in both events: 

  1. Set a safe and fun space;
  2. Introduce the concept through a starter activity (with a significant amount of scaffolding to get people used to data and movement);
  3. Lead a group reflection and discussion to help us craft future workshops;
  4. Run a longer activity involving more complex data;
  5. Close with another round of feedback.

Increasing the dataset complexity gradually was key to enhance the sense-making capacity of people. 

The first workshop was a pilot of the concept with a friendly audience. How do our ideas about data dancing play out with real people? What do we need to do to create a positive environment for collective creativity around movement and data exploration? How can we prompt movement that could represent data? What kind of research questions might we be able to answer through these activities?  

In order to find some answers to these initial questions, we created a set of movement prompts, based on a study of literature about movement and dance:

Some of the cards we created as “movement prompts”

We paired these with playful data handouts to get people choreographing short dances representing simple survey data. We were happy with how quickly this got people moving and engaged with the data. However, when we pivoted to more “real” data, like a multi-page handout about food security, many of the performances ended up looking more like silent skits. People acted out their perception of the experiences of people in the data. This was fascinating, and we think was driven by the fact that some participants didn’t have much experience  on dancing. Also, the instruction was to tell a story based on the data, so it kind of made sense people felt that acting was appropriate. If we unpack this a little more, we see that people did things like act out the hunger one felt when not having enough money to pay for food. This suggests that the data was interpreted in a very personal way when we asked people to perform it. We think people humanized the aggregate data as part of their process of acting it out. This is a fascinating potential insight; that the act of performing data could force you to relate to the lived experience of the people the data tries to represent.

Intrigued by these two key initial results, we decided to move forward in two directions so we could explore both:

  1. Work with dancers to run a second workshop focused on movement and expression;
  2. Work with a local community group to flesh out an approach to making skits with data.

The next workshop we ran focused on the first of these pathways forward – digging into the dancing aspect of our concept. What would experienced dancers do with these prompts? What ways to express data through dance hadn’t been thought of? What kinds of language and activities is this community fluent in? At that workshop, dances around the first data handout helped us explore the boundaries of the data movement cards and began to show us what kind of movement variables were good for understanding each kind of data. Participants waltzed at changing speeds to show ice cream consumption over time and moved at different scales to show happiness in Somerville. Though our first activity succeeded in focusing on dance as expression, again our second dataset led to more story-telling based movement than the first activity. We think this is a reflection of how much the data framing matters even for audiences that have a very specific skill set. Specifically, our second data handouts included a number of quotes and personal narratives as well as tables and graphs that may bring to mind the lived experience of the data and suggest a narrative format in the movement. These storylines may have been especially dominant since the workshop participants were in different dance communities with different norms and it takes time and practice to create a collective movement language and tell a story choreographically. We were grateful that the dancers shared their experiences and perspectives to help us refine our movement cards to speak the language of movement and see new opportunities for growing this approach in dancing communities. 

Both sets of workshop participants/co-designers had so much to offer! Their input and feedback and new ideas really shaped our perception of the project.

Next Steps

Like most nascent projects, our initial prototypes left us with more questions than answers or insights. Here are some of the ones at the top of our minds:

  • What more can we do with dance and data? We haven’t found the boundary yet.
  • What does data theater and physical data storytelling look like? How can we learn from theater experts?
  • What happens when we use more personal datasets or datasets with more emotional resonance and social importance? How can we approach dancing or acting sensitive datasets?
  • What does data dancing and data theater mean for society? What is it good for? Can it help new people engage with data? Can it increase empathy? Can it bring back the human side of datasets that can be dehumanizing and abstracted from reality? Can it connect data to potential paths of social action related to the issues represented in the data? Does it increase memory or understanding?

As you can see, we have a lot to explore! Concretely, that means we want to host a series of data movement workshops with a community group, focused on a dataset they’d like to explore in a new way. This will let us try to answer some of the questions about data skits and empathy. We’re also hoping to find funding to support data theater workshops with people experienced with theater; perhaps culminating in a data dance performance. We’re also hoping to host more data dancing workshops with dance groups that are already accustomed to working together and can help us find the boundaries of that space.

Do you know someone exploring similar ideas? Are you a dancer, performer, or community group who wants to look at data in a new way? Drop us a line!

Thanks to everyone who came to the workshops! Special thanks to Birgit Penzestadler and Pedro Reynolds-Cuellar who helped us to facilitate the first workshop and reflect upon our observations.  Also, thanks to the MIT Media Lab for allowing us to host these workshops in their venue.