This Quantitative Information, it vibrates?

January 15, 2005 11:20 PM

A chapter from Edward Tufte's upcoming book is online.[link contains roughly 2.2 MB of scanned images] Tufte, discussed here previously and author of what could be called the Strunk and White for scientists, statisticians, producers and consumers of visual information, takes a stab at a few issues right up the average MeFite's alley: the 9/11 commission report, fraudulent medical studies, and the rather dubious quantitative work of this unfortunate economist/art historian. For the ShillFilter suspicious, check out some of the great threads that haunt his site.
posted by fatllama (24 comments total)

Weird, I was just considering buying Envisioning Information through amazon today. I think I'll have to get it now. Thanks for this post.
posted by btwillig at 11:38 PM on January 15, 2005

Tufte is great. If you ever get a chance to go to one of his day-long seminars, by all means, do it.
posted by enrevanche at 11:43 PM on January 15, 2005

Good read, I think I'm going to buy the book.

Also liked the new word I learned: Economisting
posted by sour cream at 12:20 AM on January 16, 2005

Don't miss the pseudo-feud between Tufte and David Byrne.

That Galenson book about art production actually looks pretty interesting. It's too bad that it turns out that it might be founded on bad data and argumentation. (Although, I think I object to its thesis for entirely different reasons than Tufte gives, so maybe I should be pleased that the data doesn't run against me).
posted by painquale at 1:06 AM on January 16, 2005

One of my favourite threads is the sparklines one.
posted by dhruva at 1:09 AM on January 16, 2005

I have and love every one of his books. They don't serve me explicitly in any professional manner, but reading them is sometimes akin to folding a beautiful piece of origami. I didn't design it, but the simple beauty of it's operation is laid bare. Tufte and John Montroll are two of my adolescent favorites. Thanks for reminding me that he's still around.
posted by Jack Karaoke at 1:24 AM on January 16, 2005

re: painquale's Tufte/Byrne feud - Examples of Byrne's PowerPoint Art. Oh, no. Ouch.
posted by taz at 3:26 AM on January 16, 2005

"...few things are more appalling than listening the inept arguments and rhetorical ploys of one's allies."

This is precisely what pissed me off about Supersize Me and Fahr 9/11.
posted by srboisvert at 5:37 AM on January 16, 2005

make that:

"listening to the inept"
posted by srboisvert at 5:41 AM on January 16, 2005

If it were "the Strunk and White for scientists," Id be impressed. As it is, I don't see anything in the excerpt that addresses the horrible abuses of language perpetrated by scientific writers. They are second only to lawyers in making text unreadable.
posted by Kirth Gerson at 5:45 AM on January 16, 2005

For me, the S&W for scientists has already been written. NASA SP7010, aka Clarity in Technical Reporting by Dr. Samuel Katzoff [PDF], manages to say more about technical writing in 30 pages than most books ten times the size. There’s not a single illustration or bullet point, yet it’s lucid and precise.

Oh, and Kirth Gerson, lawyers don't do that impenetrable writing thing for fun. It's all about the job protection thing. I remember seeing SRI demo a legal document parser and indenter (like tidy for legalese). It made even pretty impenetrable stuff easy to follow. Guess it'll never make open source ...
posted by scruss at 6:14 AM on January 16, 2005

I like it, but I found the first chapter incomplete. Granted I don't read books very often and it's probably because the first chapter serves as an introduction, but I would have liked to see a more in-depth discussion of the "rhetorical ploys" he mentions in the first paragraph. I haven't found an accessible guide to rhetorical fallacies (as they are usually called) and it would be wonderful if I could pick this up and find this inside. Has anyone read the whole thing; or, is there a table of contents online?
posted by aphelion at 6:20 AM on January 16, 2005

kirth_gerson, while fatllama's phrasing might leave a bit to be desired, I'm sure that he meant it's the equivalent of a S&W for visual information.

Not sure I agree, there. Visual Display of Quantitative Information is absolutely a classic, and there's tons of stuff to learn from it about visual presentation of data. But I'm not sure that most scientists and engineers are prepared to learn those things from it just by reading the book.

I'm not sure that you could create the equivalent of S&W for grammar, FWIW. Visual "grammar" is only loosely analogous to linguistic grammar; that metaphor has been stretched far too thin for far too long, to the point where people think there actually are direct analogies from visual to linguistic "grammars".
posted by lodurr at 8:27 AM on January 16, 2005

[should have been "equivalent of S&W for visual grammar"]
posted by lodurr at 8:36 AM on January 16, 2005

Interesting that he's switching from how you present things to what you're presenting. I'm going to pick this up--he's always an eye-opener.
posted by amberglow at 8:40 AM on January 16, 2005

I think that one of the reasons why the debate between Tufte and Byrne is really in non-debate is because they are focused on two very different things. Byrne's art is focusing on the ways in which PowerPoint as a medium can be stretched to new types of design. Tufte's essay seemed to focus on the ways in which the default settings for PowerPoint encouraged design decisions that are incompatible with the presentation of certain types of scientific data.
posted by KirkJobSluder at 10:04 AM on January 16, 2005

Will this book prevent any future space shuttle disasters?
posted by gimonca at 11:40 AM on January 16, 2005

So, um, why do some of these old-school information usability guys have such crappy websites? Doesn't that dimish his genius just a bit to some people?
posted by chasing at 12:43 PM on January 16, 2005

aphelion, more draft material from the book is available, but I can't find a table of contents. Then again, he's still writing.

As it is, I don't see anything in the excerpt that addresses the horrible abuses of language perpetrated by scientific writers.

Kirth G. is right. There are more appropriate guides for scientific writing than Tufte (e.g. scruss's great link, other books). However, an increasingly important medium for communication between scientists and the public is visual. Even if the writing in JAMA were ten times better, I'd argue most Americans would get the latest medical news via the same terrible graphics and bullet-points presented in the evening news. For this reason, I suspect, Tufte is a great advocate of honesty and elegance in visual communication.

Horrible abuse of style in scientific writing is beyond common, it's the expected norm. However, I've come to believe that good science is often accompanied by good writing (though in practice this is extremely difficult even for the best intentioned researches). This was one of the best lessons to come out of the Alan Sokal Social Text affair; as Steven Weinberg puts it (emphasis mine):

I suppose that it might be argued that articles in physics journals are also incomprehensible to the uninitiated. But physicists are forced to use a technical language, the language of mathematics. Within this limitation, we try to be clear, and when we fail we do not expect our readers to confuse obscurity with profundity. It never was true that only a dozen people could understand Einstein's papers on general relativity, but if it had been true, it would have been a failure of Einstein's, not a mark of his brilliance. The papers of Edward Witten, which are today consistently among the most significant in the promising field of string theory, are notably easier for a physicist to read than most other work in string theory. In contrast, Derrida and other postmoderns do not seem to be saying anything that requires a special technical language, and they do not seem to be trying very hard to be clear. But those who admire such writings presumably would not have been embarrassed by Sokal's quotations from them.

lodur's point is interesting: I'm not sure that you could create the equivalent of S&W for [visual] grammar. I think Tufte fights this battle on two fronts. To him, a visual graphic must be honest to even be grammatical. Honesty isn't a grammatical feature in any popular natural language, but some think it ought to be. Secondarily, he strives for good style in compactness, a high data/ink ratio, and creative display of multivariate information.

To make this discussion concrete, consider the passive voice in the context of scientific writing. Tufte and many others argue that the passive voice allows a lazy author to describe effects without causes (see main link). Common side effects also include longer, complicated sentences and a dull voice; still, many scientists I know strongly prefer the passive voice because it is, in their words, dispassionate. I think there is an equivalent notion of the passive voice for visual information which is inferior stylistically for the same reasons. Bullet-point logic, trend-forecasts, and the promotion of graphical correlation to causation are all examples of visual tricks which are arguably similar to the passive voice style. Color me interested.
posted by fatllama at 12:45 PM on January 16, 2005

Aphelion, for "an accessible guide to rhetorical fallacies," try A Rulebook for Arguments, by Anthony Weston.
posted by undecided at 6:16 PM on January 16, 2005

When I first encountered Tufte's work a decade ago I was thrilled until I realized that much of what he promotes is suitable only for very high-resolution applications (print).
posted by bz at 6:25 PM on January 16, 2005

So, um, why do some of these old-school information usability guys have such crappy websites? Doesn't that dimish his genius just a bit to some people?

Well, tuft's website is certanly better then your website.
posted by delmoi at 12:49 PM on January 17, 2005

excuse me, "tufte's".
posted by delmoi at 12:50 PM on January 17, 2005

*fwds NASA SP7010 to every consultant that sends drivel to my department* thanks scruss
posted by fullysic at 5:37 PM on January 17, 2005