Home » Carrie Roy: When Art Collides With Data at TEDxMileHighWomen (Transcript)

Carrie Roy: When Art Collides With Data at TEDxMileHighWomen (Transcript)

Here is the full transcript of Carrie Roy’s TEDx Talk: When Art Collides With Data at TEDxMileHighWomen conference. This event occurred on October 28, 2016.


I am a data artist. I work to transform numbers into art, and that art helps people engage with information.

I work with experts in many different fields, and that is awesome. But sometimes, the collaborative process can begin like a bad joke. A journalist and a data artist walk into a bar. And the journalist says she wants to convey how important it is that massive amounts of dairy farm manure stays out of our drinking water. At this point, I slide my water to the side; I am just drinking beer.

She says in Brown County, Wisconsin, there’s on average a half cow for every acre of agricultural land. Now I think in pictures, so I’m envisioning this landscape dotted with half-Holsteins. But I wonder, how much waste does each produce? I look it up: 65 pounds of manure in one day.

And I remember thinking, yes, that’s a lot of manure! But I’m not exactly sure how much; it’s hard to visualize. Now, maybe it was the beer, but the numbers sparked my imagination.

What if I had people stand next to it? Like this life-sized half-cow standing on whatever 65 pounds of manure is. Now, as an artist, this is a big job. A life-sized half-cow is a challenge, and when you have a big challenge, you want to use your best tools. So, I used my computer, three saws, an angle grinder, and my vision as a sculptor; and Ms Brown, as we call her, emerged from the wood chips and went on to tour seven cities, raising awareness on water quality issues.

We have data visualization down to a science, when I would argue some data require more humanity, a reflection on the memories, emotions, and sensory experiences that make us human. Graphs and charts are great for some data, but we have an increasing need to integrate many types of data to enable new insights.

So I’m going to touch on the following three data challenges: grabbing attention, making the abstract tangible, and tackling complexity. And here too, as humans, we need to use our best tools to address these challenges. So number one: I could draw on a map where the threatened sage grouse lives, or I could show you hundreds of images where it lives and help us reflect on the human impact on its habitat.

ALSO READ:   Jamie Mason Cohen: How to Spot a Leader in Their Handwriting (Transcript)

I could tell you that one in three wells in the state of Wisconsin has detectable levels of pesticide or herbicide, or evoke your memory of approaching a faucet in this sculpture, the middle third being black walnut, a wood that’s toxic when you first cut into it. When we are awash in information every single day, these powerful tools – our memories, emotions, and senses – can help us pause to think more deeply about an issue we decide to care about.

Related to the challenge of grabbing attention is number two: making the abstract tangible 1,275 people responded to the survey prompt, “What is a fake piece of hair worn by a man?” Most said toupee, wig, hairpiece, rug, and all their responses are represented by different colors and patterns here.

Now, sure, only one person said, “cootie garage,” one person said, “SIB, Some is Bought” My favorite, “sky piece,” I celebrated with the comb-over-inspired pattern you see at the bottom. Now you see it, yes.

Typical charts don’t let you explore the statistically insignificant, but in this case, they were hilarious, and they help highlight our human tool of humor and how enjoyable it can be to use. What words were used only by female Victorian authors, or only by male Victorian authors? I could show you the list, or I could show you.

Well, here are three words from the list. ‘Comfits’ is from the female list. Comfits is a sugar-coated Victorian treat. Or, I could transform the results through 3-D printing into a Victorian inkwell, a metaphor that illuminates the different wells of words these men or women drew from and help us reflect on why. How Southern was William Faulkner? 83 per cent, based on running Faulkner’s works against Southern words from the Dictionary of American Regional English. Now most surprising was just how bad a computer was at identifying Southern regionalisms: only about five per cent correct.

The human component is celebrated in this work: a collection of his works in a book, when opened to 83 per cent, reveals a portrait of the author. This art and research walks the line between culture and data in a way that challenges us. I see my art as sketches in how we relate to information, but they also fuel my imagination regarding how we can engage with information in the future.

ALSO READ:   3 Tools to Become More Creative: Balder Onarheim (Transcript)

And this brings us to number three: complexity, a challenge requiring our most advanced tools. In my field of digital humanities, we take the most beautiful works of human expression and turn them into numbers to help us answer questions. For example, Charles Dickens died halfway through writing The Mystery of Edwin Drood. We analyzed all of Dickens’ other works in an attempt to predict as much as we could about the missing chapters. Everyone wants to know who did it, right?

Well we had a steamer trunk full of data, word length, character count, gender analysis, and much more, but very little that would satisfy a mystery lover. Now maybe, if we could have integrated more of these bits and pieces of data, it could’ve brought us a bit closer, but it’s a big challenge. A common challenge today – multiple streams of data, related, but separate, and we are missing the big, compelling story.

In healthcare, how can we help children with type 1 diabetes, and their families, make sense of three sources of data? So, carbohydrates – those are their meals – with insulin doses help keep their blood glucose levels from getting too high or too low. Well, humans are uniquely playful, another great tool, so I sought inspiration from games.

I combined the data into one landscape. So, clouds are carbohydrates, the landscape is an area graph of blood glucose, with red at dangerously high or low levels, and then a stream of insulin dosing runs beneath. All of the numbers are visually represented here, and the goal is simple: a rolling green landscape.

By combining the data into one landscape, it’s easier to integrate the three types of data, and we can see how one affects another. In a calendar setting, families can make sense of daily or weekly trends. I grew up reading physical landscapes, trapping gophers in North Dakota. Animals leave a wealth of data in their environment. In this picture, the feathers and the animal tracks tell a story of a bird grabbing its prey and lifting off.

ALSO READ:   Staffan Ehde: Who Decides What You Think? Not You... at TEDxYouth@Helsingborg (Transcript)

About a decade ago, I interviewed an old trapper, and he said something that caught my attention: “I would only set traps for male mink”. And I said, “How could you tell that they were male?” He said their tracks were slightly larger. He went on to note that their tracks could indicate an animal’s maturity, how fast they were going, where they were likely coming from or going to based on food or water in the area, clues on timing of movement and more.

We are all here today because our ancestors mastered this tool: the ability to read all of these points of information, data, in a 3-D environment and synthesize the information to form a full, compelling story.

Now, what if that animal were financial markets or human health? These examples are also part of complex data landscapes. While we’re good at identifying important data, we tend to separate it out, separate graphs, and then we miss the big picture. Can you imagine if we did that with animal tracks? I mean, sure, we could identify the species, but all those other details I mentioned would be lost without context. And mink don’t leave symbols or colors on an X-Y axis. This takes mental energy to keep straight. We can interpret data left by the mink because we know what a mink looks like, moves like, and the ground.

The physical world is our framework for interpretation. But we could really use any familiar framework to help us engage with data. I’ve described data in landscapes, but they exist on objects as well. For generations, master stoneworkers or woodworkers have carefully read 3-D information on the surface of a stone or a piece of wood, and they rely on that information to precisely cut, strike, or cleave that unique raw material into their vision, and their art is a testament to this remarkable human tool. One obvious challenge is we’ve never been able to create high resolution data objects or data landscapes quickly or inexpensively.

Pages: First |1 | ... | | Last | View Full Transcript