The Thinking Eye

24 Oct 2016

Data Expert

At a recent talk that Edward Tufte gave at Art Center College of Design in Pasadena, CA, he presented some of his new material. Those familiar with his books (and if you’re not, you should really check them out) will appreciate the glimpse at what this master of visualization is currently working on.

In true Tufte fashion, the talk’s program was beautifully crafted. Here’s page one and page two.

Here are a few notes I jotted down from his talk. They reference what it takes to be a Thinking Eye, a discerning observer and decipherer of the way the world is.

Galileo was one of the best thinking eyes ever. Of course, he had a lot of new things to see. His notebooks are full of knowledge-packed diagrams. One minute with Galileo’s notebooks is worth 100 minutes of #datavis.

Thinking eyes see and then thinking eyes act. To produce, to execute, to act, is essential to the process.

For whom does the thinking eye perform? For ones self.

Thinking Eyes are of this world, they are empirical, they are practical.

Tufte went on to describe in detail 6 characteristics of the Thinking Eye.

  1. See verbs and interactions. Tim Berners-Lee’s original vision of the internet included statement about moving from a hierarchy of nouns to a flat landscape of verbs (links and hypertext).
  2. See now, words later. See to learn, not just to confirm. See with “vacation eyes”, as if for the first time. For example, having a description of an art piece greatly colors its perception.
  3. Focus and serenity. Serenity is a condition under which all brainpower is devoted to The Thinking Eye. In doing creative work, do not start you day with addictive time vampires such as the NYT, email, etc. Instead begin right away with your work.
  4. A sense of the relevant. Define what is important in a pile of data or knowledge. Scan globally, focus locally. In learning about the world, you will rarely learn more by any method other than going out into the field and observing how the data are generated. Those numbers on the screen are often representation of representations. Go see the real interactions in the world, and lose the fog of data.
  5. Recognizing excellence. Play in the big leagues and do things that have a long time horizon, if possible. The distribution of excellence is fiercely lognormal, skewed, long-tailed. Excellence provides standards. As an exercise, place any visualization next to Google Maps and cry. Google Maps has layers upon layers of information without becoming cluttered, and it is the most viewed visualization in the history of the mankind.
  6. Understanding the relationship between information and conclusions, evidence and inference. “How do you know that?” The most cited article in the medical sciences in the past 10 years is “All published studies are false”. The study of human behavior isn’t rocket science – it’s harder than rocket science. You can have your own point of view, but not your own facts. If you torture the data long enough, it will confess to anything.
comments powered by Disqus