Nathan, over the FlowingData blog, points to this video where John Tukey himself discusses the analysis of multivariate data using computers… in 1972. The library contains other great videos, so I encourage you to explore it.

Tukey had an enormous impact in the way we look at the data, but exactly who are “we”? Are “we” the readers of the NYT that can’t make sense of a scatter plot? Can “we”, proud of our technology do, using an ordinary tool (say, Excel), what a dinosaur computer was able to do 40 year ago? Do “we” know what a boxplot is?

A friend of Tukey, Tufte, tells us that “John Tukey once mentioned that when he wanted some graphic inspiration, he would read Bertin”.


Jacques Bertin, French cartographer, the other founding father of modern graphical analysis, still vastly ignored by Anglo-Saxon world outside the academia (no English Wikipedia entry, writing it is one of my new year’s resolution). He published his seminal work, Semiologie Graphique, in 1967 (translated to English in 1983…). I love this quote:

It is the internal mobility of the image which characterizes modern graphics. A graphic is no longer ‘drawn’ once and for all; it is ‘constructed’ and reconstructed (manipulated) until all the relationships which lie within it have been perceived.

One must look at the images in his books to see what “internal mobility” meant, with the available technology in 1968.

His work is complex (and his writing style doesn’t help) but I’ll try to write some posts during 2008 to show how bright his work is and how we can use it to better understand our data.