Edward Tufte, reigning guru of information design, believes that someday "real soon now" interfaces will be as seamless as they used to be – 400 years ago.
Open a book. A page of typewritten text can hold more than 5,000 characters. Look at a map. If it's reasonably well designed, a single square inch of a page-sized map can reveal a dozen or more pieces of information, as well as many relationships.
Now turn on your computer. You're navigating through complex information aided by a state-of-the-art graphical user interface, but you're back in the dark ages of information resolution: Even the most precise display screens work at data densities only about one-tenth of those of a map or a page of text. The latest
versions of high-end word processors feature "full-screen" functions in which all menus and icons are hidden and people can just write. They still don't present as much information as paper.
And that bothers Edward Tufte, the reigning guru of information design. Built around classics of information presentation taking in everyone from Christopher Scheiner (both a contemporary and a rival of Galileo) to Mafia attorney Bruce Cutler, Tufte's books The Visual Display of Quantitative Information (1983) and Envisioning Information (1990) emphasize how the presentation of information can enhance the data being manipulated. In recent years, Tufte has investigated how computer screens can present more and better information. He has done some corporate consulting (he refers to one client as a "certain international business machine company in Armonk, New York") and, along with a colleague at Yale, has devised a method for updating traditional hospital-patient records. During semester breaks, he offers one-day courses that elaborate on the ideas introduced in his books.
Tufte calls computers "just about the lowest-resolution information interface known to humankind, compared to the map, the photograph, printed text, or the brain. You have to do special things in a low-resolution world: not waste pixels, use anti-aliased typography, give all the space to the user, because it's so precious."
Tufte then suggests the heretical idea that GUIs aren't perfect. "Icons are pretty inefficient," he says when I present him with the opening screen of Norton Desktop for Windows. "They give you one piece of information, just one noun, and they waste 4,000 pixels on the icon, so you get 50 choices about where to go next. If this screen used text, you could probably get 500 choices. You get much deeper to where you want to be with one choice. It's an interesting trade-off: You have a decorative 50 choices, or you have a flat, clunky 500 choices. The problem for people using a low-resolution environment is that we have to go through a number of steps to get to where we want to be and we don't get much of what we want when we get there. The problem for designers is that they have to announce to the world, 'Look at me!' " And we get pretty screens with minimal information.
For the record, Tufte is a Mac man. He waxes enthusiastic whenever the conversation turns to the designers at Xerox PARC or those who came up with the Macintosh interface, and his sprawling home and office are filled with more than a half-dozen Macs and large monitors.
"The Mac interface is most of the way there," states, but then suggests that it's limited by the paradigm of operation systems and applications. "Either we get rid of applications or we get an application with enough universality that you have the capacity of Illustrator, Photoshop, and Quark all together. You can then forget applications and focus on documents or objects. The idea is to expose most of the capabilities right up front, so people know what tools are available most of the time. That exposure has to be done with incredible efficiency and tightness so that it doesn't eat up the interface with administrative debris."
As Tufte sees it, today's GUIs are problematic, in part, "because we've been taught to segregate text and graphics. There are a couple sources for that fallacy. One of them is that the modes of production have segregated text and image and tables up until recently. The other is that some people have taken a physiological speculation about how the brain works and tried to turn it into a design strategy: all that left brain/right brain, text/image nonsense. The funniest thing about the left brain/right brain book on drawing is that they have the drawings of children and they show what the children drew before and afterward, and the befores are much better than the afterwards. The afters are sort of stereotypical; the befores are idiosyncratic and quirky."
As he does throughout his books and his lectures, Tufte looks to history for design solutions. "Someday – as they say in the industry, 'real soon now' – interfaces will be as seamless as (they were for) Galileo." He opens an original edition of Galileo's Istoria e dimostrazioni intorno alle macchie solari e loro accidenti, then turns to a description of the first telescopic observations of Saturn. In the text itself, as if it were a word, is a picture of Saturn and its rings. The integration is seamless. "So 'real soon now,' we'll be able to do what Galileo did in 1613."
Applications nowadays are stuffed with features and mini-applications, enough so that people are often intimidated by the hidden features and don't bother learning them. "We've all heard jokes about programming VCRs, but the same is true of the features of any of the main word processing programs. There are lots of things that users would like to do, but the learning costs are too high. There's a joke along the lines of 'if a million monkeys had a million years they could type out the complete Shakespeare.' Similarly, if a million monkeys had a million years at a terminal they could eventually get two columns in Word. The key thing in judging an interface – and this is how the trade press should review things – is not the features but the access to the features. Sure, the features exist, but do they exist as far as most users are concerned?"
Tufte sounds like he's preaching when he prescribes a design solution. "Usually there isn't much evidence for a great design strategy. A good design has a wholeness, an integrity, and a oneness that comes from a single intelligence. That's true of all great designs. They're idiosyncratic, they have a personality. It's hard to give evidence why one design is better than another. I try to, that's what my work is. I try to give it some principles, but it's a difficult question. A great design really has the quality of revelation."
