Learning to See

I took a portrait drawing class many years ago. The most important lesson of all, which I still find myself thinking about on a regular basis today, nearly 20 years later, is that in order to draw, you need to re-learn how to see.

Specifically, you need to learn to see light and shadow.

Normal people don't think about light and shadow when looking at a face. They think about features, like noses, eyes, and hair. But when trying to recreate that face, you need to mostly concentrate on the light and shadow, because that's actually how you see things like noses.

You don't really "see" the object, you see light reflecting off of that object. And the position of the highlight on one side, and the shadow on the other side, is what tells your brain there's a protrusion of a certain shape and size that we understand to be a "nose".

One of the first portraits I did that I was proud of was based on the cover of Miles Davis' album "Tutu".

The photo this is based on was very dramatically lit and allowed me to easily see the light and shadows, which might otherwise have escaped my untrained eyes. I was learning to see differently.

In data visualization, we have a similar need. But the requirements shift a bit. Rather than thinking about light and shadow, we need to think about our data, the number and nature of variables we're communicating, and how to allow our audience to "see" that information.

What we need to do, is translate that information into visual "marks and channels", as Tamara Munzner refers to them in her book Visualization Analysis and Design. I interviewed her last week for an upcoming episode for my Lesson and Listen series and it got me thinking (for the umpteenth time) about how we learn to see data attributes, and specifically how to think of that vision happening pre-attentively (subconsciously and very quickly), which critical to understand in data visualization design.

But how do we learn to see in a new way? We practice. Just like I took a class in portrait drawing, and stared at photos and learned to detect subtle shifts in light and shadow and use that to decide where and how to apply charcoal to a piece of paper, so do we need to practice, look at others' creations of marks and channels and think about how they're the translations of data meant to serve a specific purpose.

Simply put: work on your "seeing" skills and you'll get better at your visualizing skills.

==========

Ask me about virtual or in-person data storytelling and visualization training for your team

Learn more about data storytelling and visualization via my other LinkedIn Learning courses.