Reading 1: Edward Tufte: Envisioning Information

The chapter I decided to read was “Layering and Separation”. In this chapter, Tufte discusses how the most important idea when layering various data together, is that the relationship between these layers is defined and there is “proportion and harmony”. I thought the idea of macro-annotation of data was very interesting, as there are some situations where there is simply too much data that needs to be displayed, and the annotations turn the data into an easily digestible “story”. When he’s discussing dark vs light grid lines in blank sheet music paper, he brings up the idea of “chartjunk”. Continuing further in the article, it seems that Tufte isn’t planning on focusing explicitly on the idea but diving into the concept of “1 + 1 = 3”, where extra unnecessary activity appears if you’re not careful how you structure organize the colors and thickness of various lines in your design. Overall, his argument is obviously “don’t clutter your information”, and that you can avoid this by being very deliberate with your lines.

Reading 2: Stephen Few: The Chartjunk Debate

Few starts out his argument by reminding us of the definition of chartjunk, which was important because it never came up in the section that I read. “According to Tufte, chartjunk consists of non-data and redundant data elements in a graph. It comes in various types: sometimes artistic decoration, but more often in the form of conventional graphical elements that are unnecessary in that they add no value”. Few is trying to argue that while the basic foundation of Tufte’s argument of chartjunk is sound, he believes that there are some useful cases for redundant embellishments and that they sometimes are even more helpful that just minimalist charts. He backs this up with a recent study done by the University of Saskatchewan. However he then proceeds to dive into the faults of the study. One that particularly stood out to me was that the plain graphs in the study were actually quite ugly, and went against many of the points Tufte brought up in the “Layering and Separation” chapter I read. He also narrowly tailors Tufte’s definition of chartjunk, to emphasize that nothing that supports the chart’s information is junk, which I think is quite important.