The Lost Art of Hand-made Graphs; Or, The Rise of Mass Produced Charts

I recently read this post by Dr. Drang over at And now it's all this on the lost art of making charts1. This article reminded me of something that is blindingly obvious in retrospect: until very recently, almost all the portrayals of scientific data were hand-drawn.

Take this graphic of Napoleon's attempted invasion of Russia by Charles Joseph Minard2. Or Dr. John Snow's map identifying the locations of cholera outbreaks in London. Both were done by hand, and both are very artful.

And now we have these. And these. And slightly better, these.

Graphics have become industrialized. You can tell whether or not someone is using R, Matlab, Excel, or gnuplot by inspection. Different communities have different standards. But they're definitely no longer done by hand. Sometimes this is good. Sometimes this is bad.

On the opposite end of the spectrum, just thirty odd years ago, everyone had to do mathematical typesetting by hand. And now we have Donald Knuth's gift to the world: LaTeX3. Though that can also make papers like these seem way more legit than they should4.

It's interesting to see scientific analysis become more mass produced.

  1. Now these things are called 'infographics' or 'data visualization.' I don't really know what makes those things different from charts.

  2. Apparently Minard is considered by some the Father of Infographics.

  3. Not to mention his contribution to the world via The Art of Computer Programming.

  4. I have to admit that when I read a paper on the arXiv typeset in Microsoft Word, I automatically deduct ten to twenty legitimacy points. I shouldn't, but I do.