Review: Beautiful Evidence by Edward Tufte
Permalinks to this entry: individual page or in monthly context. For more material from my journal, visit my home page or the archive.
Bloggers and particularly web and print designers consider Edward Tufte a guru. He's a professor emeritus at Yale, a sculptor, and perhaps the world's foremost authority on how to be clear and accurate when presenting complex information visually. Beautiful Evidence is his latest book. It is a lovely book, a manifesto, a screed, very personal, and also, in its own elegant way, a bit of a mess.
The book is full of brilliant ideas and strategically repeated concepts. He has two overarching messages:
- That human beings are very good at understanding and interpreting remarkably dense and complicated representations of information, from large statistical tables to vast and precise visual maps, so we deserve to see that kind of detailed data analysis rather than the facile and trivial "chartjunk" so often generated by PowerPoint and other low-resolution tools.
- That visual representations work best when treated as an integral part of every piece of work, so that diagrams, photos, charts, tables, and graphs should appear within the text they're part of, rather than abstracted away because of the false limitations of printing or display technology.
The book is also a lovely piece of design that takes its own advice, ensuring that the many example photos, charts, tables, and diagrams sit right where they should, near the text that discusses them—sometimes even being repeated in several locations so you never have to flip back and forth to find what Tufte is talking about. I learned a ton.
On the other hand, Beautiful Evidence is surprisingly slapdash. I don't mean that it wasn't prepared carefully—I found only one very minor typo, which is unusually few for a book of its size, and it is printed and bound meticulously—but that it feels less like a cohesive work than a series of semi-related essays assembled in almost-random order.
Tufte is fixated on the integrated text-and-image approaches of Galileo and Richard Feynman, who are great examples, to be sure, but sometimes I felt like he was hitting me over the head with them—yes, yes, Edward, I get that they were both visually and intellectually brilliant and we should follow them, now find me someone else too! He inserts his own jargon—two-space, three-space, chartjunk, Phluff, economisting—as if everyone understood it, which seems contrary to the message he's preaching.
And why, suddenly, at the end of the book, do we get a long quoted interview about the techniques of a Shah-era Iranian protester on his techniques for pulling down statues with ropes, followed by several full-spread photos (some oddly blurry) of Tufte's own large sculptures? Why, indeed, is the dust cover a montage of four action photos of a dog jumping into a pond? If there's an explanation, I missed it.
These are quibbles, mostly. Tufte's wonderful arguments compensate for them. But I can't help thinking that Tufte could have used a more involved editor—the book is essentially self-published—to help him come across as less arrogant. His contempt for poor information design is refreshing, but I think it could have been organized a little better.
Still, Beautiful Evidence is a great and important book. It's a spectacular deal too, only $58 Cdn from Amazon.ca (I get a cut if you buy it from my link) or $52 USD from Tufte's site directly, postage included. (That's less than I paid for a black-and-white, softcover calculus textbook in 1986.) I don't know how he can afford to print and sell such a lush hardcover volume for such a low price.