Revisiting the Language of Data Visualization: 3 Themes to do Viz Better

Years ago, in a blog far away, I started thinking through how to do visualization better. Better, while it sounds tangible, is about as abstract as black holes. What counts as success?

I knew what I liked and couldn’t do.

I knew people that seemed like they had a handle it.

I could look at Twitter, Tableau Public, and why yes, Upshot, Financial Times, and data driven journalism at large and know those cats were crushing it.

Then, I’d look at my own work, and something was missing.

I started this blog back then as a way to push boundaries in my experience. Now, I’m writing a book. If you’ve ever written anything longer than 13 pages, you know this takes an absurd amount of time. Here’s what I’ve learned some 250 pages later.

Charts aren’t words, paragraphs, or pictures

As we look to familiar cliches to put charts into a meaningful context, finding that exact match is hard. Charts are bigger than a word, sometimes smaller (or larger) than a paragraph, and aren’t an exact lens into what we see in the world like photos.

American Sign Language uses classifiers to provide details about position, movement, shape, and a host of other details. (No, this isn’t in the book…yet…) They are not signs (or words) in the traditional sense, but a system to which meaning is applied. (The system part is.) In ASL, I can talk about a car driving through the hills. I’ll tell you about the cars and hills and – with a classifier (just one) – I can show how it moves across the rough terrain. At times, I can convey paragraphs worth of information with a single classifier, or it can be a single sentence worth of information. The value and effectiveness varies widely.

English has classifiers, but they’re not nearly as exciting as ASL classifiers. In a number of spoken languages, classifiers end up being boring part words like a sheet of paper or a gaggle of geese. These words end up having specific matches and are terrible to learn as second language learners. Cookies can be made in and discussed in batches, but steak is not. Sticks can be bundles, but flowers are bouquets. Annoyed yet?

Charts are systems to which we apply meaning. Sometimes, that meaning adds up to a 1,000 words. Other times, they’re barely worth half a sentence. Charts are like ASL classifiers. They need other things – like titles, dimensional words, and of course numbers – to be interesting.

Why this matters:

• When we see charts as systems, this changes our approach. Certain charts (and classifiers) go with certain concepts, but there’s freedom in how we implement them.
• Classifers aren’t good or bad, but tools in making communication more specific. No, they cannot stand alone, but need guardrails. You decide how many and where.
• We can use one or many charts to achieve the desired effect and to land the specific nuance we hope to convey.

Graphicacy is a 3rd tier skill

Eons ago (or so it seems), we used to discuss the 3 R’s: reading, writing, and arithmatic. Fundamentally, these 3 skills formed the basis of what we needed to operate in a modern society.

Folks, that is changing rapidly. Technology shifts feed into culture, reshaping the skills needed to function in society. Try operating in today’s society without some type of device (computer, cell phone, etc) and it will be difficult. Most jobs require an online application. Digital skills – whether by phone, tablet or computer – are a key part of today’s interactions with the world.

But, interfaces aren’t the only thing changing. How we document the world is also transforming. Both static and moving images play a pivotal role in our communication. Video and audio are a core part of how we make information meaningful. And charts are becoming a larger part of how we communicate information. Where numeracy helps us navigate arithmetic and literacy folds in both reading and writing, graphicacy explores how we interpret and create charts.

The story of the Covid-19 pandemic is one told in charts. Financial Times – and specifically John Burn-Murdoch – is among several documenting everything about the pandemic with charts taking the lead. “Flatten the curve” relies on an understanding that the curve is created by data from cases.

Graphicacy – and its closely related cousin data literacy – is part of the every day lexicon. We are experiencing the sociological shift people felt when reading didn’t just become common, but a basis for learning and understanding society. People reference stories in shorthand: “turning into a pumpkin” relies on knowledge of the story Cinderella and that it’s time to go. Literacy is a part of our lexicon and informs how we understand the world. We “read” people, just like a book. Charts are starting to make that leap with “flatten the curve.”

Yet, we are far from the Data Enlightenment. Tableau has commited to an enormously large spend to close the gap on data literacy skills. Be Data Lit, an initiative from Allen Hillery and Sarah Nell-Rodriquez, approaches and highlights communities typically ignored in this space: underrepresented groups, reskilling career changes, and so many others not typically centered in the story around data literacy.

We have a whole chapter on data literacy.

How this affects you, the analyst:

• When we think about graphicacy like reading, we become a lot more empathetic to the leaps people have to make. We can begin to bridge the gap on why line charts might be hard.
• We can also recognize where and why people might have gaps in knowledge. A great book fills in gaps a variety of ways, whether by clarifying, providing another means of understanding like illustrations, or giving references to learn more.
• We can fight for broader access. A truly literate culture means everyone has a set level of skills to navigate it.

Not just visualization, but communication

As data visualization continues to move forward, the medium itself is expanding. It has to, honestly, to reach the most amount of people. While virtualization of life in general comes through a screen, we interact with it in a myraid of ways. Sometimes, I can touch and gesture on the screen, giving me a sense of something tactile. People are playing with sonification, playdoh, and other mediums in which to share data. This increases accessibility and fosters a multimodal environment. Data physicalization isn’t new. Khipus are a 500 year-old technology logging everything from social rank, tax payment, and occupation.

But we’re also moving data visualization from a standalone medium to one intermixed everywhere. Charts appear in newspapers as the lead, they’re a part of our discussions as we pull up information on phones to settle debates, but perhaps most key to this point, they’re moving into social mediums like Slack to be a part of the conversation. Tools like Data Wrapper exist to quickly share charts almost completely on the fly. We need data and charts to have effective conversations.

Beyond seeing the data, we need mediums that allow us to personalize the shape of data to the modes we best understand. Audiobooks expand access to the written word. Braille and raised letters physicalize a visual medium, letting fingers scan information quickly and silently. Slack and chatboxes are just beginning.

The nuts and bolts:

• Emily Kund made a stellar point about accessibility and innovation. Replace “accessibility” with “innovation” and see how you react. Widening access makes a better product.
• We’re not monomodal creatures. Think about typing. You feel the keys, your body has potentially learned where each key lives, and there’s potentially a voice in your head as you write. The sounds of the keys clatter in different ways and your eyes may gaze down.
• We’re all different. Some cats absolutely love to be read to. I was one of those kids I just wanted to steal the book away and read it for myself. Little has changed in that arena, but I have a kid that loves to be read to, so I’m the one reading now.

---

Coda

Last year at this time, we snagged a book deal and began thinking about the book we’d write. Some things we expected, most things we did not, and we’re approaching the end of writing soon. Which means we need to start marketing. Did you like this post? We have a whole site dedicated to this idea and will be starting a newsletter soon with similar content, better grammar, and maybe even some videos.