I completely agree with the author, but I don't think it was taken far enough. The positive examples at the bottom are crawling out of whatever pit of hell the "184 milk jugs" came from, but meaningless colors, exuberant gradients, fading bars, big overlapping data points... these are all examples of chart junk as well.<p>My favorite information design mantra is that your design should be <i>exactly</i> as exciting as your data. This probably takes it too far, but much like statistics it is so easy to lie with graphs that you're better off assuming you're lying accidentally.
FYI, the link to the Cleveland article is broken. It should read<p><a href="https://secure.cs.uvic.ca/twiki/pub/Research/Chisel/ComputationalAestheticsProject/cleveland.pdf" rel="nofollow">https://secure.cs.uvic.ca/twiki/pub/Research/Chisel/Computat...</a>