Q&A: Julie Steele

In your talk, you mentioned beauty and elegance applied to data visualisation. How does this apply in science? 

The definition of beauty that I referenced was an elegant solution to some kind of problem. And science is about solving problems, but half of the battle is stating the problem well—having a good question you are trying to answer. So the “elegant solution” part refers to a solution that is as simple as possible, but still answers the question.

The solution can sometimes be creative, although creativity is not required. The best example is the periodic table; sometimes things stand the test of time and there is no need to reinvent the wheel. I think the other thing that contributes to a beautiful solution is being able to understand how you arrive at it. If nobody can replicate your solution and if no one else can understand the thinking that got you there, then it’s not elegant and it’s not simple.

In scientific journals, we often see the same formatting pattern for all kinds of data visualisation. What do you think about those patterns?

Conventions are useful. Expectations are important so that people come in expecting a certain thing, and when you meet those expectations, you are making it easier for them to connect with your ideas because you are speaking a language that they understand. However, I agree that sometimes it’s important to break the conventions either to drive people out of a certain way of thinking and to pay more attention or when the convention is not working anymore.

It’s important to find the right tools for the right job. It’s important to fit the format of whatever you are doing, whether it’s writing or visualisation, to your content.

And it’s important to let the content dictate what it needs in order to be well  presented. In visualisation, we have structures such as the x-axis and the y-axis, which we label with numbers and draw lines. Or we use tables and pie charts because we know they work well in a given situation. But if a piece of data requires something new because it won’t fit in any other, then it’s OK to change the structures. Consistency is valuable because it communicates and meets expectation, which makes communication efficient. But, I don’t think that consistency is the primary value; it’s not so sacred that we can’t touch it.

Why did you become an editor? 

I fell in love with Shakespeare, which I’m sure is the typical English major answer! I knew from the time I was fifteen that I loved words and language and that this was a way to connect with other human beings. That was important to me. I happen to have majored in British literature, but I’ve always been fascinated by the way the world is put together. And I always wanted to poke at it and find holes in it and put it together in a different way, which is what programmers do all day long; I just chose a different form of language.

That’s part of why I also enjoy political science, because it is about building a structure or a system and poking at it and writing new laws to patch the holes. And it’s also why I enjoy publishing books about computer science because it’s all about system design and finding the ways we humans are irrational and surprising in the way we interact with systems that are designed to be robust